UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A systematic review of the blood pressure lowering efficacy of calcium channel blockers in the treatment… Wong, Michelle Mon Yee 2007

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2007-0284.pdf [ 11.75MB ]
Metadata
JSON: 831-1.0100803.json
JSON-LD: 831-1.0100803-ld.json
RDF/XML (Pretty): 831-1.0100803-rdf.xml
RDF/JSON: 831-1.0100803-rdf.json
Turtle: 831-1.0100803-turtle.txt
N-Triples: 831-1.0100803-rdf-ntriples.txt
Original Record: 831-1.0100803-source.json
Full Text
831-1.0100803-fulltext.txt
Citation
831-1.0100803.ris

Full Text

A SYSTEMATIC REVIEW OF T H E B L O O D PRESSURE L O W E R I N G EFFICACY OF CALCIUM C H A N N E L BLOCKERS IN T H E T R E A T M E N T OF PRIMARY > HYPERTENSION  by MICHELLE MON YEE WONG  A T H E S I S S U B M I T T E D IN P A R T I A L F U L F I L L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E OF M A S T E R OF S C I E N C E  in  T H E F A C U L T Y OF G R A D U A T E STUDIES  (Pharmacology and Therapeutics)  T H E U N I V E R S I T Y OF BRITISH C O L U M B I A March 2007 © Michelle Mon Yee Wong, 2007  ABSTRACT Context - Calcium channel blockers (CCBs) are widely used drugs to lower elevated blood pressure and manage angina and arrhythmias. Although the goal of antihypertensive therapy is to lower the risk of cardiovascular disease-related morbidity and mortality, efficacy is most often gauged by blood pressure reduction.  Objectives — This systematic review of the blood pressure lowering efficacy of CCBs aims to determine the dose-related changes in systolic blood pressure (SBP), diastolic blood pressure (DBP), heart rate, and withdrawals due to adverse events ( W D A E ) with C C B treatment compared with placebo for a duration of 3-12 weeks, in patients with primary hypertension (SBP > 140 m m H g and/or D B P > 90 m m Hg).  Design — A systematic review, as per the methodology of the Cochrane Collaboration, of randomized placebo-controlled trials.  Methods - Electronic databases were searched using a modified, expanded version of the search strategy used by the Cochrane Hypertension Review Group. RevMan 4.2 software was used to analyze data.  Participants — 106 trials were included and reported data on 13 878 patients with a mean age of 55 years, mean baseline blood pressure of 158.2/101.6 m m H g , mean pulse pressure of 56.7 mm H g , and mean treatment duration of 5.7 weeks.  Results — Maximal blood pressure lowering efficacy of CCBs is achieved at twice the manufacturer's recommended starring doses. dihydropyridines and 8/6  This maximal reduction is 10/7 m m H g for  m m H g for non-dihydropyridines and likely represents  an  overestimate of the true blood pressure lowering effect due to publication bias. Combined,  ii  dihydropyridines and non-dihydropyridines lower pulse pressure by 3 m m H g (95% CI: -4, 2). Compared with placebo, W D A E increased in a dose-related fashion for dihydropyridines [relative risk of 1.8 (95% C I 1.2, 2.6) at 2 times the starting dose compared with 3.9 (95% CI: 2.2, 7.0) at 4 times the starting dose]. There were insufficient data to make a conclusion about the effect of non-dihydropyridines on W D A E .  Conclusion — Dihydropyridines reduce blood pressure to a greater degree than nondihydropyridines. Maximal blood pressure lowering for both subclasses occurs with twice the manufacturer-recommended starring dose.  Increasing the doses of dihydropyridines  above recommended starting doses increases withdrawals due to adverse effects.  iii  TABLE OF CONTENTS  ABSTRACT  ii  TABLE OF CONTENTS  iv  LIST O F T A B L E S  ix  LIST O F F I G U R E S  xii  LIST O F A B B R E V I A T I O N S  xiri  PREFACE  xiv  ACKNOWLEDGEMENTS 1. B A C K G R O U N D I N F O R M A T I O N 1.1 H Y P E R T E N S I O N  1.1.1 Definition of hypertension and its relation to cardiovascular morbidity and mortality • 1.1.2 Blood pressure variability, pulse pressure and heart rate: relation to cardiovascular outcomes 1.1.3 Types of chronic hypertension 1.1.3.1 Isolated systolic hypertension 1.1.3.2 Primary hypertension 1.1.3.3 Secondary hypertension 1.1.4 Management and treatment of chronic hypertension 1.1.4.1 Non-pharmacological treatment 1.1.4.2 Pharmacological treatment 1.1.5 Blood pressure reduction and cardiovascular events 1.1.6 Blood pressure measurement 1.1.7 Dose-response relationships 1.2 T H E R O L E O F C A L C I U M I N T H E C A R D I O V A S C U L A R S Y S T E M  1.2.1 Calcium 1.2.2 Calcium Channels in the cardiovascular system 1.3 C A L C I U M C H A N N E L B L O C K E R S  1.3.1 Historical Aspects 1.3.2 Classification .1.3.3 Pharmacodynamics of Calcium Channel Blockers 1.3.3.1 Molecular mechanism of action 1.3.3.2 Physiological effects 1.3.3.3 Phenylalkylamines 1.3.3.4 Benzothiazepines  xv 1 1  1 3 4 4 5 6 6 6 7 9 10 12 12  12 13 14  14 15 16 16 17 18 18  iv  1.3.3.5 Dihydropyridines 1.3.3.6 Benzimidazolyls 1.3.4 Pharmacokinetics of calcium channel blockers 1.3.5 Clinical use of calcium channel blockers 1.4 S Y S T E M A T I C R E V I E W S  1.4.1 1.4.2 1.4.3 1.4.4  What are systematic reviews? H o w does a systematic review differ from a narrative review? The Cochrane Collaboration Arm of this systematic review  2 .PROTOCOL  19 19 20 21 27  27 29 30 30 32  2.1 O B J E C T I V E S 32 2.2 M E T H O D O L O G Y 32 2.2.1 Types of studies 32 2.2.1.1 Why are only randomized controlled trials included? 33 2.2.1.2 Why is a parallel placebo arm necessary? 33 2.2.1.3 Why is blinding (masking) necessary? 34 2.2.1.4 Why is a baseline measurement subsequent to or during a washout/placebo run-in period important? 35 2.2.1.5 Why is the 3-12 week window selected? 35 2.2.2 Types of participants 35 2.2.3 Types of interventions '. 36 2.2.4 Types of outcome measures 36 2.2.5 Search strategy for identification of studies 37 2.2.6 Study selection 42 2.2.7 Data extraction 42 2.2.8 Quality assessment 44 2.2.8.1 The Cochrane approach for assessment of allocation concealment.. 45 2.2.8.2 Jadad method 46 2.2.9 Data analysis 46 2.2.10 Statistical considerations 49 2.2.10.1 Individual study data 49 2.2.10.2 Pooling trials 50 2.2.11 Starting doses 51 2.2.12 Direct and indirect comparisons between doses 52 3. R E S U L T S 3.1 S E A R C H F I N D I N G S 3.2 C H A R A C T E R I S T I C S O F I N C L U D E D S T U D I E S 3.3 C H A R A C T E R I S T I C S O F S T U D I E S A W A I T I N G D A T A F R O M A U T H O R S 3.4 C H A R A C T E R I S T I C S O F E X C L U D E D S T U D I E S 3.5 O V E R V I E W O F I N C L U D E D T R I A L S 3.6 V A L U E S U S E D T O I M P U T E M I S S I N G V A R I A N C E S  54 54 55 134 136 146 148  v  3.6.1 Standard deviation of blood pressure change 148 3.6.2 Standard deviation of heart rate change 149 3.7 D O S E - R E L A T E D B L O O D P R E S S U R E L O W E R f N G O F I N D I V I D U A L C C B DRUGS 149 3.7.1 Amlodipine vs. placebo 150 3.7.2 Barnidipine vs. placebo 153 3.7.3 Darodipine vs. placebo 154 3.7.4 Felodipine vs. placebo 155 3.7.5 Isradipine vs. placebo 159 3.7.6 Lacidipine vs. placebo 162 3.7.7 Lercamdipine vs. placebo 163 3.7.8 Manidipine vs. placebo '. 165 3.7.9 Nicardipine vs. placebo 166 3.7.10 Nifedipine vs. placebo 168 3.7.11 Nilvadipine vs. placebo 170 3.7.12 Nisoldipine vs. placebo 171 3.7.13 Nitrendipine vs. placebo 172 3.7.14 Prandipine vs. placebo 174 3.7.15 Summary of blood pressure lowering efficacy of dihydropyridines 175 3.7.15.1 Dihydropyridines - Assessment of publication bias 176 3.7.15.1.1 Dihydropyridines -subgroup analysis based on trial size 176 3.7.15.1.2 Dihydropyridines -trim and fill method 177 3.7.16 Diltiazem vs. placebo 180 3.7.17 Verapamil vs. placebo 182 3.7.18 Tiaparnil vs. placebo 184 3.7.19 Summary of blood pressure lowering efficacy of non-dihydropyridines 185 3.7.19.1 Non-dihydropyridines - Assessment of publication bias 186 3.7.19.1.1 Non-cuhydropyridines-subgroup analysis based on trial size... 186 3.7.20 Other CCBs 188 3.7.20.1 Lidoflazine vs. placebo 188 3.7.20.2 Mibefradil vs. placebo .189 3.8. B L O O D PRESSURE VARIABILITY  3.8.1 3.8.2 3.8.3 3.8.4  Systolic vs. diastolic blood pressure Calcium channel blockers vs. placebo Systolic vs. diastolic blood pressure entry criteria Baseline vs. endpoint  3.9. P U L S E PRESSURE 3.10. D O S E - R E L A T E D C H A N G E I N HEART RATE O F INDIVIDUAL C C B DRUGS  3.10.1 Dihydropyridines vs. placebo 3.10.1.1 Amlodipine vs. placebo 3.10.1.2 Darodipine vs. placebo 3.10.1.3 Felodipine vs. placebo 3.10.1.4 Isradipine vs. placebo 3.10.1.5 Lercanidipine vs. placebo 3.10.1.6 Manidipine vs. placebo 3.10.1.7 Nicardipine vs. placebo 3.10.1.8 Nifedipine vs. placebo 3.10.1.9 Nisoldipine vs. placebo  190  190 191 191 192 193 :.. 194  194 195 196 196 197 197 198 198 199 199  3.10.1.10 Nitrendipine vs. placebo 3.10.2 Non-dihydropyridines vs. placebo 3.10.2.1 Diltiazem vs. placebo 3.10.2.2 Verapamil vs. placebo 3.10.2.3 Tiapamil vs. placebo 3.10.3 Other calcium channel blockers 3.10.3.1 Lidoflazine vs. placebo 3.10.3.2 Mibefradil vs. placebo  200 200 201 202 202 203 203 203  ,  3.11 DOSE-RELATED WITHDRAWALS DUE TO ADVERSE EVENTS  205  3.11.1. Dihydropyridines vs. placebo 3.11.1.1 Amlodipine vs. placebo 3.11.1.2 Darodipine vs. placebo 3.11.1.3 Felodipine vs. placebo 3.11.1.4 Isradipine vs. placebo 3.11.1.5 Lercanidipine vs. placebo 3.11.1.6 Manidipme vs. placebo 3.11.1.7 Nicardipine vs. placebo 3.11.1.8 Nifedipine vs. placebo 3.11.1.9 Nilvadipine vs. placebo 3.11.1.10 Nisoldipme vs. placebo 3.11.1.11 Nitrendipine vs. placebo 3.11.1.12 Prandipine vs. placebo 3.11.2. Non-dihydropyridines vs. placebo 3.11.2.1 Diltiazem vs. placebo 3.11.2.2 Verapamil vs. placebo 3.11.3. Other calcium channel blockers 3.11.3.1 Mibefradil vs. placebo  205 206 206 207 207 208 208 209 210 210 211 211 212 212 213 214 215 215  . DISCUSSION 4.1  216  WHAT METHODOLOGICAL ISSUES A N D POTENTIAL SOURCES ENCOUNTERED WHILE CONDUCTING T H E SYSTEMATIC REVIEW?  4.1.1 Publication bias 4.1.2 Selection bias  OF BIAS WERE 216  ,  218 220  4.2  WHAT IS T H E DOSE-RELATED BLOOD PRESSURE LOWERING EFFICACY OF EACH SUBCLASS? IS THERE A DIFFERENCE IN T H E BEST ESTIMATE OF T H E MAGNITUDE OF B P LOWERING EFFECT OF DIFFERENT SUBCLASSES OF CCBS? 221  4.3  IS THERE A DIFFERENCE IN T H E BEST ESTIMATE OF T H E MAGNITUDE OF B P LOWERING EFFECT OF DRUGS IN EACH SUBCLASS? 222  4.4  WHAT IS T H E EFFECT ON BLOOD PRESSURE IN T H E PLACEBO GROUP IN SHORT-TERM TRIALS? 222  4.5  D O E S T H E METHOD OF BLOOD PRESSURE MEASUREMENT AFFECT T H E BLOOD PRESSURE LOWERING EFFICACY OF CCBS? 223  4.6  D O E S TRIAL QUALITY AFFECT T H E BLOOD PRESSURE-LOWERING CALCIUM CHANNEL BLOCKERS?  4.7  WAS THERE A DIFFERENCE IN BLOOD PRESSURE LOWERING EFFICACY AT TROUGH VS. PEAK? ...224  EFFICACY OF 223  vii  4.8  DID  FUNDING  SOURCE AFFECT  THE  REPORTED  BLOOD  PRESSURE  LOWERING  EFFICACY OF C C B S ?  224  4.9  D O E S A G E A F F E C T T H E B L O O D PRESSURE L O W E R I N G EFFICACY O F C C B S ?  225  4.10  D O E S CO-MORBIDITY A L T E R T H E B L O O D PRESSURE L O W E R I N G EFFICACY O F  226  CCBs? 4.11  DOES BLOOD  PRESSURE L O W E R I N G  EFFICACY OF  CCBs  DIFFER  FOR  ISOLATED  S Y S T O L I C H Y P E R T E N S I O N VS. D I A S T O L I C O R S Y S T O - D I A S T O L I C H Y P E R T E N S I O N ? . . . . 2 2 6 4.12 H o w  D O T H E DIRECT COMPARISONS B E T W E E N DOSES DIFFER F R O M T H E  INDIRECT  COMPARISONS? 4.13 F O R E A C H CCB  227 DRUG, D O T H E MANUFACTURER'S STARTING DOSES COINCIDE  T H E L O W E S T E F F E C T I V E D O S E AS D E T E R M I N E D B Y THIS SYSTEMATIC R E V I E W ? 4 . 1 4 W H A T IS T H E E F F E C T O F 4 . 1 5 W H A T IS T H E E F F E C T O F  CCBs CCBs  ON  BP  229  O N PULSE PRESSURE?  231 WITH RESPECT TO  C H A N G E IN HEART RATE?  232  4 . 1 7 IS T H E R E A N Y E V I D E N C E O F A D O S E - R E P O N S E R E L A T I O N S H I P  WITH RESPECT TO  WITHDRAWALS D U E TO ADVERSE EVENTS? CAN  THE MAGNITUDE  OF BLOOD  232  PRESSURE L O W E R I N G  EFFICACY OF  C H A N N E L BLOCKERS BE LINKED TO THEIR MECHANISM OF ACTION? 4.19  HOW  C A N TFIE  BLOOD  228  VARIABILITY?  4 . 1 6 Is T F I E R E A N Y E V I D E N C E O F A D O S E - R E S P O N S E R E L A T I O N S H I P  4.18  WITH  PRESSURE L O W E R I N G  EFFICACY OF CALCIUM  B L O C K E R S IN S H O R T - T E R M TRIALS B E R E L A T E D T O T H E I R E F F E C T S O N A N D M O R B I D I T Y O U T C O M E S IN L O N G - T E R M TRIALS?  CALCIUM 233 CHANNEL  MORTALITY 234  5. C L I N I C A L I M P L I C A T I O N S  237  6. I M P L I C A T I O N S F O R F U T U R E R E S E A R C H  240  7. R E F E R E N C E S  243  8. A P P E N D I C E S  277  APPENDIX A - TRIAL SELECTION FORM  277  A P P E N D I X B - STANDARD D A T A E X T R A C T I O N FORM FOR E A C H TRIAL  278  viii  LIST O F T A B L E S Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table  1: Classes of current drugs used to treat hypertension 7 2: Comparison of 3 hypertension guidelines 8 3: Classification of group A calcium antagonists 16 4: Group B calcium antagonists 16 5: Pharmacokinetic parameters of CCBs 21 6: Long term clinical trials of CCBs for treatment of hypertension 24 7: Data input for continuous data in RevMan 49 8: Data input for dichotomous data in RevMan 49 9: Formulae for individual study responses for dichotomous data in RevMan 49 f 0: Starting doses of calcium channel blockers 52 11: Results of search strategy 54 12: Amlodipine — Characteristics of included studies 56 13: Barnidipine — Characteristics of included studies 62 14: Darodipine — Characteristics of included studies 63 15: Diltiazem — Characteristics of included studies ; 64 16: Felodipine — Characteristics of included studies 73 17: Isradipine — Characteristics of included studies 81 18: Lacidipine — Characteristics of included studies 89 19: Lercanidipine — Characteristics of included studies 90 20: Lidoflazine — Characteristics of included studies 94 21: Manidipine — Characteristics of included studies 95 22: Mibefradil - Characteristics of included studies 97 23: Nicardipine— Characteristics of included studies 100 24: Nifedipine — Characteristics of included studies 105 25: Nilvadipine — Characteristics of included studies 113 26: Nisoldipine — Characteristics of included studies 115 27: Nitrendipine — Characteristics of included studies 116 28: Pranidipine — Characteristics of included studies 123 29: Tiapamil — Characteristics of included studies : 124 30: Verapamil - Characteristics of included studies 125 31: Characteristics of studies awaiting office blood pressure data from authors ... 134 32: Reasons for exclusion of certain studies meeting inclusion criteria 136 33: Overview of included trials using CCBs as monotherapy for primary hypertension 146 34: Blood pressure lowering efficacy of amlodipine 1.25-10 mg/day 150 35: Blood pressure lowering efficacy of barnidipine 10-30 mg/day 153 36: Blood pressure lowering efficacy of darodipine 100-300 mg/day 154 37: Blood pressure lowering efficacy of felodipine 2.5-20 mg/day 155 38: Blood pressure lowering efficacy of felodipine in older and younger subgroups. 156 39: Blood pressure lowering efficacy of felodipine in trials with inclusion criteria of systolic/systodiastolic hypertension vs. diastolic hypertension 157 40: Blood pressure lowering efficacy of isradipine 1-20 mg/day 159 41: Blood pressure lowering efficacy of isradipine according to formulation 160 42: Blood pressure lowering efficacy of lacidipine 2-4 mg/day 162 43: Blood pressure lowering efficacy of lercanidipine 2.5-20 mg/day 163 44: Blood pressure lowering efficacy of manidipine 10-40 mg/day 165  ix  Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table  45: Blood pressure lowering efficacy of nicardipine 40-120 mg/day 166 46: Blood pressure lowering efficacy of nifedipine 20-100 mg/day 168 47: Blood pressure lowering efficacy of nilvadipine 8-30 mg/day 170 48: Blood pressure lowering efficacy of nisoldipine 10-30 mg/day 171 49: Blood pressure lowering efficacy of nitrendipine 5-20 mg/day 172 50: Blood pressure lowering efficacy of pranidipine 1-8 mg/day 174 51: Summary of blood pressure lowering efficacy of dihydropyridines 175 52: Blood pressure lowering efficacy of dihydropyridines combined according to multiples of starting dose 176 53: Dihydropyridines: post-hoc subgroup analysis of trough B P lowering based on trial size 177 54: Blood pressure lowering efficacy of diltiazem 90 - 540 mg/day 180 55: Blood pressure lowering efficacy of verapamil 60 — 540 mg/day 182 56: Blood pressure lowering efficacy of tiapamil 300-1200 mg/day 184 57: Summary of blood pressure lowering efficacy of non-dihydropyridines 185 58: Blood pressure lowering efficacy for non-chhydropyridines combined according to multiples of starting dose 186 59: Non-dihydropyridines: post-hoc subgroup analysis of trough blood pressure lowering based on trial size 187 60: Blood pressure lowering efficacy of lidoflazine 180 mg/day 188 61: Blood pressure lowering efficacy of mibefradil 6.25-100 mg/day 189 62: Variability of SBP and D B P at end of treatment '. 190 63: Baseline standard deviations of blood pressure according to entry criteria 191 64: Standard deviations of B P at baseline vs. endpoint in trials with D B P entry criteria 192 65: Change in pulse pressure 193 66: Effect of dihydropyridines on heart rate 195 67: Effect of amlodipine on heart rate 195 68: Effect of darodipine on heart rate 196 69: Effect of felodipine on heart rate 196 70: Effect of isradipine on heart rate 197 71: Effect of lercanidipine on heart rate 197 72: Effect of manidipine on heart rate 198 73: Effect of nicardipine on heart rate 198 74: Effect of nifedipine on heart rate 199 75: Effect of nisoldipine on heart rate 199 76: Effect of nitrendipine on heart rate 200 77: Effect of non-cuhydropyridines (diltiazem and verapamil) on heart rate 201 78: Effect of diltiazem on heart rate 201 79: Effect of verapamil on heart rate 202 80: Effect of tiapamil on heart rate 202 81: Effect of lidoflazine on heart rate 203 82: Effect of mibefradil on heart rate 203 83: Effect of dihydropyridines on withdrawals due to adverse events 205 84: Effect of amlodipine on withdrawals due to adverse events 206 85: Effect of darodipine on withdrawals due to adverse events 206 86: Effect of felodipine on withdrawals due to adverse events 207 87: Effect of isradipine on withdrawals due to adverse events 207  Table Table Table Table Table Table Table Table Table Table Table Table Table  88: Effect of lercarridipine on withdrawals due to adverse events 208 89: Effect of manidipine on withdrawals due to adverse events 208 90: Effect of nicardipine on withdrawals due to adverse events 209 91: Effect of nifedipine on withdrawals due to adverse events 210 92: Effect of nilvadipine on withdrawals due to adverse events 210 93: Effect of nisoldipine on withdrawals due to adverse events 211 94: Effect of nitrendipine on withdrawals due to adverse events 211 95: Effect of pranidipine on withdrawals due to adverse events 212 96: Effect of non-dihydropyridines on withdrawals due to adverse events 213 97: Effect of diltiazem on withdrawals due to adverse events 213 98: Effect of verapamil on withdrawals due to adverse events 214 99: Effect of mibefradil on withdrawals due to adverse events 215 100: Comparison of manufacturer's recommended starting doses and lowest effective doses determined in this systematic review : 228  xi  LIST OF FIGURES Figure Figure Figure Figure  1: Binding sites of calcium channel blocker drugs 17 2: Example of a forest plot 47 3: L o g dose-response curve for amlodipine 1.25-10 mg/day 152 4: Funnel plot of standard error against effect size of change in SBP for amlodipine 5 to 20 mg/day 152 Figure 5: Funnel plot of standard error against effect size of change in SBP for felodipine 5 to 20 mg/day 158 Figure 6: Funnel plot of change in SBP for dihydropyridines at maximal blood pressure lowering 178 Figure 7: Funnel plot of change in D B P for dihydropyridines at maximal blood pressure lowering 179  xii  LIST O F A B B R E V I A T I O N S  Ambulatory Blood Pressure Monitoring Blood Pressure Calcium Channel Blocker Calcium ion Cardiovascular Cochrane Collaboration Confidence Interval Congestive heart failure Consolidated Standards of Reporting Controlled Delivery Coronary heart disease Diastolic Blood Pressure Double-Blind Electrocardiogram Female Gastro-fntestinal Therapeutic System Hazard Ratio Heart Rate fsolated Systolic Hypertension Joint National Committee Male ' Mean Arterial Pressure Medical Subject Heading Mercury Milligram Millimetres of mercury Multicentre Odds Ratio Placebo-Controlled Randomized Controlled Trial Randomized Relative Risk (Risk Ratio) Review Manager Slow-Release Standard Deviation Standard Error of the Mean Systolic Blood Pressure Weighted Mean Difference Withdrawals Due to Adverse Events World Flealth Organization - International Society of Hypertension  ABPM BP CCB Ca CV CC CI CHF CONSORT CD CFID DBP DB ECG f GfTS - HazR HR 1SH JNC m MAP MeSH Hg mg mm H g MC OR PC RCT R RR RevMan SR SD SEM SBP WMD WDAE WHO/fSH 2+  PREFACE "...evidence based medicine requires you not only to read papers but to read the right papers at the right time and then to alter your behaviour (and, what is often more difficult, the behaviour of other people) in the light of what you have found." - Trisha Greenhalgh, 1997 (from: Flow to read a paper: the basics of evidence based medicine. London: B M J Publishing Group, p. 2)  xiv  ACKNOWLEDGEMENTS I wish to thank my supervisor, D r . James M . Wright, for his patience and guidance. His dedication to revealing the truth about drug efficacy and safety has inspired me to be involved in evidence-based medicine research. I would also like to thank my other supervisory committee members, D r . David Godin, D r . Casey V a n Breemen and D r . Michael Walker, for their comments and expertise. I wish to express my gratitude to all the members of the Therapeutics Initiative and the Cochrane Hypertension Group for their support over the last few years. I wish to thank Dr. K e n Bassett and Dr. T o m Perry, Jr. for their guidance on my critical appraisal projects and Mr. Ciprian Jauca for his coordinating efforts for the T I and the Hypertension Group. I would like to extend my appreciation to those who helped me execute this systematic review. Especially, I am indebted to D r . Vijaya Musini for her assistance and direction throughout every stage of this review. I am grateful to M r . Benji Heran for being the second independent reviewer and for developing a new, high-yield search strategy. M r . Stephen Adams retrieved a myriad of articles for this systematic review, and his efficiency was invaluable. I also wish to thank D r . Sonia Franciosi, Ms. Laurence Jacquaz, D r . Marco Perez, D r . Frederique Rodieux, D r . Stephan Schwarz, D r . Michelle V a n den Engh, and D r . Alexander Zolotoy for translation of foreign language articles. I must also thank all of the clinical trial authors who took the time and effort to correspond with me during the process of developing this review. I would like to thank my systematic review colleagues, of past and present, for their camaraderie: Benji, Marco, Jenny, Vijaya, Cremona, D a l and Jose. Last but not least, I thank my family for their unconditional support in all my endeavors.  xv  1. B A C K G R O U N D I N F O R M A T I O N  1.1 Hypertension Hypertension, or elevated blood pressure, depending on how it is denned, is present in up to 20-30% of adults (1). Although often referred to as a "silent disease", hypertension is actually a surrogate marker and major risk factor for cardiovascular disease (2), rather than a disease itself. Elevated blood pressure is a risk factor for stroke, coronary artery disease (CAD) and congestive heart failure (CHF), and these cardiovascular diseases are the most common causes of morbidity and mortality in developed countries. Elevated blood pressure has been estimated to cause 4.5% of global disease burden (3).  However, developing  countries have a greater share of the global burden of cardiovascular disease mortality than developed countries (4).  1.1.1.  D e f i n i t i o n of hypertension and its relation to cardiovascular morbidity and mortality The Joint National Committee on Prevention, Detection, Evaluation and Treatment  of High Blood Pressure (JNC) describes the relationship between systolic blood pressure and cardiac risk as "strong, continuous, graded, consistent, independent, predictive and etiologically significant" (5).  Rather • than using a strict numerical divide between  normotension and hypertension, it is more practical to define hypertension as "that level of blood pressure above which investigation and treatment do more good than harm" (6), or stated more precisely by Kaplan: "that level of blood pressure at which the benefits (minus the risks and costs) of action exceed the risks and costs (minus the benefits) of inaction" (7). With this pragmatic clinical approach, the definition is subject to change as new evidence  1  regarding the treatment of hypertension becomes available. This definition also underscores the importance of individualizing the diagnosis and treatment of hypertension. There are two views of the epidemiologic data relating B P to adverse cardiovascular events.  One view is that the risk is linear and continuous and the other is that there is a  threshold. The concept of individualizing the diagnosis of hypertension according to age and sex was the focus of a recent reanalysis of epidemiological data from the Framingham Heart study (8).  Using a logistic spline model, they proposed that overall mortality is  unrelated to SBP at pressures below an age- and sex-dependent threshold (approximately at the 70  th  percentile).  However, above the 80  th  percentile, the mortality risk increases  logistically. The numerical criteria used to define normotension and hypertension are thus arbitrary and subject to change. In another study involving a meta-analysis of observational data from individual patients aged 40-89 years, they found a direct relationship between vascular mortality (due to stroke, ischemic heart disease and other vascular casuses) and blood pressure for each decade of age (9). For ages 40-69, each difference of 20 m m H g from the "usual" SBP for each age group, or 10 m m H g from the "usual" D B P age was associated with a two-fold difference in vascular mortality (9). However, there were no data representing patients with blood pressures less than 115/75 mm H g , and data on overall mortality were also not reported. Although elevated blood pressure itself is generally asymptomatic, the target organ damage resulting from chronic hypertension is the antecedent to cardiovascular morbidity and mortality. In hypertension, the increased afterload on the heart leads to diastolic and systolic dysfunction, which can subsequently lead to heart failure.  Combined with the  increase in myocardial oxygen demand, the decrease in coronary oxygen supply that results from the hypertension-induced acceleration of atherosclerosis can lead to myocardial  2  infarction. Arterial damage from hypertension also contributes to the development of aortic aneurysm/dissection, stroke, retinopathy, nephrosclerosis and renal failure. According to 1999 Canadian Hypertension Guidelines, chronic hypertension is diagnosed in 3 clinic visits i f resting B P exceeds the arbitrary threshold of 140/90 m m H g and there is target organ damage or resting B P is over 180/105 m m H g , and in 5 visits over 6 months i f these features are absent but the resting mean B P remains over 140/90 m m H g (10).  1.1.2.  B l o o d pressure variability, pulse pressure and heart rate: relation to cardiovascular outcomes  The degree of variation of blood pressure throughout a 24-hour period is important clinically because it correlates with severity of target organ damage and contributes to the risk of cardiovascular mortality in hypertensive patients (11).  Blood pressure variability  tends to be increased in patients with primary hypertension (11). Furthermore, increased systolic B P variability has been found to predict progression of carotid artery wall atherosclerosis and rate of cardiovascular events independently of increased blood pressure itself (12).  More prospective studies utilizing continuous ambulatory blood  pressure  monitoring are necessary to confirm the prognostic value of blood pressure variability. Pulse pressure, calculated as the difference between systolic and diastolic blood pressure, can be considered an element of blood pressure variability because it reflects the blood pressure changes within the cardiac cycle (11). A marker of arterial stiffness, the magnitude of the pulse pressure also depends on stroke volume and the speed of reflected pressure waves. Pulse pressure correlates not only with surrogate outcomes such as carotid artery damage, but also with major cardiovascular endpoints. For example, clinic-measured pulse pressure has been linked to recurrent events after myocardial infarction in patients with  3  left ventricular dysfunction (13).  The Progetto Ipertensione  Umbria Monitoraggio  Ambulatoriale (PIUMA) registry of subjects with primary hypertension showed that ambulatory pulse pressure was a strong independent predictor of total cardiovascular risk, with better prognostic value than pulse pressure derived from conventional B P readings (14). A n epidemiologic study showed that pulse pressure appeared to be the best blood pressure parameter in predicting mortality in people > 65 years old (15). Elevated resting heart rate is also considered an important predictor of hypertension and a risk factor for cardiovascular and non-cardiovascular death. In the Frarningham study, those with a baseline resting heart rate of greater than 84 beats per minute were shown to be at significandy increased cardiovascular risk than those with lower heart rates (16). Blood pressure parameters  aside, many other risk factors that contribute to  cardiovascular disease have been identified through monitoring of the Frarningham study population (17). These include non-modifiable factors such as male gender, increased age (>55 years for men, > 65 years for women), positive family history of premature cardiovascular disease, prior cerebrovascular accident or myocardial infarction, and modifiable factors such as smoking, dyslipidemia, diabetes mellitus, increased weight, low physical activity and left ventricular hypertrophy.  1.1.3. Types of chronic hypertension  1.1.3.1 Isolated systolic hypertension The definition of isolated systolic hypertension (ISH) is in flux and varies from SBP > 160 m m H g and D B P < 95 mm H g to SBP > 140 m m H g and D B P < 90 m m H g (17). Since systolic blood pressure rises with age, I S H is the most common form in the elderly. Pulse pressure is increased in this population as well. The main cause of an increase in SBP  4  (and concomitant decline in D B P ) is thought to be tidckening and stiffening of conduit arteries resulting from degeneration of the arterial wall after exposure to chronic cyclic stress (18). This reduced arterial compliance enhances the amplitude and velocity of the pressure wave originating from the heart. Hence, the pressure wave is reflected from peripheral to central arteries earlier, thus increasing the pressure in late systole (18). The increases in afterload and myocardial work can lead to left ventricular hypertrophy.  Coupled with  diminished coronary perfusion pressure, these increased strains on the heart could contribute to myocardial ischemia.  1.1.3.2. Primary hypertension Primary hypertension, or so-called "essential" hypertension comprises over 90% of hypertensive patients, and implies that there is no known cause.  Vascular alterations in  hypertension include decreased lumen diameter, increased reactivity and stiffness, and increased wall thickness-to-lumen ratio of resistance vessels (19). Rather than attributing these alterations merely to vascular growth, some studies have shown that the same amount of vascular material is restructured via a process termed "eutrophic remodelling" (20). Hypertension is also associated with impairment of endothelial nitric oxide activity, though it remains to be discovered i f nitric oxide dysfunction is a cause or consequence of hypertension (21). According to the mosaic theory of primary hypertension, the interplay of multiple genetic and environmental factors contributes to elevated arterial pressure (22). Because of this multifactorial nature, there is large heterogeneity in individual responses to any antihypertensive medication. The ongoing Genetics of Hypertension Associated Treatment (GenHAT)  study  is assessing  how  hypertension  susceptibility genes interact  with  antihypertensive drugs to modify blood pressure response and risk of coronary heart disease  5  (23). This pharmacogenetic study is analyzing various polymorphisms o f genes regulating the renin-angiotensin-aldosterone system, sodium-volume homeostasis, insulin resistance, and sympathetic nervous system activation.  1.1.3.3. Secondary hypertension Hypertension that arises secondary to some identifiable mechanism is less common. Several secondary causes of hypertension have been identified, including renal (e.g. nephritis, renal artery stenosis), endocrinological (e.g. Cushing's syndrome, phaeochromocytoma), and neurological (e.g. encephalitis, porphyria).  Other causes include coarctation of the aorta,  pregnancy, and alcohol or drug use.  1.1.4. Management and treatment of primary hypertension  1.1.4.1 Non-pharmacological treatment A l l hypertension guidelines agree on the use o f non-pharmacological control of blood pressure as first-line treatment, including diet alteration (e.g. reduced intake o f salt), weight reduction, smoking cessation and dynamic exercise. A systematic review o f shortterm sodium restriction in patients with elevated blood pressure demonstrated  modest  decreases in SBP and D B P (by about 4 and 2 m m H g , respectively) (24). Another Cochrane review o f long-term restriction in dietary salt found an even smaller effect size, with decreases in SBP by 1.1 mm H g and in D B P by 0.6 m m H g (25). A n analysis o f 44 trials that assessed the effect of exercise on resting B P demonstrated that the average reductions in SBP and D B P in hypertensive patients were 7.4 m m H g and 5.8 m m H g , respectively, compared to a reduction of 2.6 m m H g and 1.8 m m H g in normotensive patients (26). Because of a lack of relationship between exercise frequency, time per session, or intensity and the magnitude of blood pressure reduction, it is suggested that the dose-response curve is flat (26).  6  1.1.4.2.  Pharmacological treatment  From a historical perspective, pharmacological treatment of hypertension has evolved from a few agents with many adverse effects that were reserved only for severe/malignant hypertension, to a wide array of currendy available drugs with diverse mechanisms of action (Table 1). During the 1940s, veratrum alkaloids, thiocyanates, and ganglion blockers such as hexamethonium were some of the only pharmacological alternatives to bilateral sympathectomy and rigid sodium restriction (27). From the 1950s onwards, a better understanding of the pathogenesis of hypertension has led to the development several new classes of antihypertensive drugs: rauwolfia alkaloids such as reserpine, vasodilators such as hydralazine, peripheral sympathetic inhibitors such as guanethidine, and thiazide diuretics such as chlorothiazide. In the 1960s, beta blockers and centrally acting sympathetic inhibitors (e.g. alpha methyldopa) were introduced.  In the  1970s, alpha adrenergic blockers and angiotensin converting enzyme inhibitors were made available, followed by calcium channel blockers in the 1980s and angiotensin II receptor blockers in the 1990s.  Table 1: Classes of current drugs used to treat hypertension Example Class Thiazide diuretics Loop diuretics Beta blockers A C E Inhibitors Alpha blockers Calcium channel blockers Autonomic agents Angiotensin II receptor antagonists  Hydrochlorothiazide Furosemide Propranolol Benazepril Prazosin Amlodipine, verapamil, diltiazem Reserpine Losartan  Currendy, the most optimal management of hypertension is unclear.  Supposedly  evidence-based guidelines are highly variable with respect to the thresholds for initiation of anti-hypertensive therapies and the choice of initial drugs (Table 2) (28).  7  T a b l e 2: C o m p a r i s o n of 3 hypertension guidelines (28) D r u g treatment B P C a n a d i a n Joint Natinional threshold Committee ( J N C ) VI  No target organ damage or risk factors With C V risk factors (other than diabetes mellitus) With target organ damage With diabetes mellitus or renal disease Choice of initial drugs < 60 years old  > 60 years old  160/100 m m H g (160/105 m m H g i f 60 yrs. or older) 160/90 m m H g  140/90 m m H g  World Health Organization/ International Society of Hypertension 1 9 9 9 150/95 m m H g  140/90 m m H g  140/90 m m H g  160/90 m m H g  130/85 m m H g  140/90 m m H g  140/90 m m H g  130/85 mm H g  130/85 m m H g  Thiazides, blockers or inhibitors Thiazides, acting CCBs  Diuretics blockers  beta ACE long-  or  beta  Thiazides, beta blocker-thiazide combinations, or long-acting CCBs  A l l available classes  drug  Diuretics or CCBs  With the numerous classes of antihypertensive drugs available, selection of a firstline agent should be based on mortality and morbidity data from long-term randomized controlled trials. Wright et al's systematic review of randomized controlled trials of first-line antihypertensive therapies selected studies of at least one year's duration that provided data on morbidity and mortality (29).  The review showed that low-dose thiazide diuretics  decreased mortality (relative risk [RR] 0.89, 95% confidence interval [CI] 0.81-0.99], stroke (RR 0.66, 95% C I 0.56-0.79), coronary artery disease (RR 0.71, 95% CI 0.60-0.84) and total cardiovascular events (RR 0.68, 95% CI 0.62-0.75) compared with placebo or no treatment. Eligh-dose diuretics showed similar results for stroke and total cardiovascular events, except there was no statistically significant difference compared with no treatment in death and  8  C A D . For first-line beta blockers there was no statistically significant difference compared with placebo/no treatment for all four outcomes.  For calcium channel blockers, one  placebo-controlled trial (SYST-EUR) showed a reduction in the risk of stroke (RR 0.61, 95% CI 0.43-0.87) and cardiovascular events (RR 0.71, 95% CI 0.57-0.87) but not death and CAD.  In terms of blood pressure lowering efficacy, the drop in SBP was statistically  significandy greater with thiazides than with beta blockers or CCBs, while the drop in D B P was similar among all three of these classes. Using  a  technique  called "network  meta-analysis",  Psaty  et  al. combined  cardiovascular endpoint data from both placebo-controlled and comparative trials of firstline antihypertensive agents (30).  For C A D , C H F , stroke, cardiovascular disease events,  cardiovascular mortality and total mortality, low-dose thiazide diuretics were superior to placebo. None of the other classes of agents (beta blockers, A C E inhibitors, C C B s , alpha blockers, angiotensin receptor blockers) were significantly better than low-dose diuretics in any of the listed outcomes. In comparison with C C B s , low dose diuretics were associated with a reduction in C H F (RR, 0.75; 95% CI, 0.67-0.81) and cardiovascular disease events (RR, 0.94; 95% CI, 0.89-1.00). Thus, results from both Wright et al. and Psaty et al.'s analyses support the use of low-dose thiazide diuretics as first choice for first-line drug therapy for treatment of hypertension. The J N C guidelines support this approach (17).  1.1.5.  B l o o d pressure reduction a n d cardiovascular events Meta-analyses  antihypertensive  of major  clinical  trials have  demonstrated  clear  benefits  therapy, including reductions in cardiac and cerebrovascular  of  events.  Although the goal of any antihypertensive therapy is to lower the risk of cardiovascular  9  disease-related morbidity and mortality outcomes (5), efficacy is most often gauged by blood pressure reduction, a validated surrogate outcome.  1.1.6.  M e t h o d s of b l o o d pressure measurement Blood pressure is a highly variable measurement that is modified by many unknown  and known factors, including circadian rhythm, seasonal variation, activity level and emotional state. Despite this obstacle, blood pressure measurement is a universal tool in medical practice.  Since hypertension is generally clinically silent, accurate diagnosis and  treatment would not be possible without proper use of instruments to measure blood pressure (31). There are two modalities of measuring blood pressure: 1) non-invasively, using an indirect method by which pressure in an occlusive cuff is correlated with blood flow phenomena, and 2) invasively, using a catheter to obtain a direct measure (32). The indirect method of auscultation is the historical standard for recording clinic blood pressure.  The Korotkoff sounds involved in indirect blood pressure measurement  originate from the intra-vessel turbulence and vibrations during compression of the brachial artery. The appearance of Korotkoff sounds (phase I) as the bladder cuff deflates correlates with the systolic blood pressure, while the disappearance of these sounds (phase V) best reflects the diastolic blood pressure.  There are several factors that can interfere with  accurate blood pressure measurement — these arise from the patient, the measurer, the instrument and the technique (33). Given the fluctuation of blood pressure during clinical measurements, the average of two or more blood pressure measurements in a single arm is a more reliable than a single reading. There are also substantial variations in blood pressure throughout the day and with routine activities. Other patient factors such as bowel or bladder distension, emotional extremes, recent ingestion of alcohol, and even talking can  10  elevate blood pressure.  Measurer factors include the rate of cuff inflation and deflation,  hearing acuity, and expectation bias, including end-digit preference. include device calibration, cuff size, and arm position.  Technique factors  The most common physician  mistakes in blood pressure measurement include using an inappropriately small cuff, failing to allow a pre-measurement rest period of 5 minutes, deflating the cuff too fast, failing to measure in both arms, and failing to palpate maximal systolic blood pressure before auscultation (33). Automated instruments utilize auscultatory, oscillometric or Doppler ultrasound techniques.  The most commonly used devices are oscillometric and employ proprietary  algorithms to determine blood pressures. Despite existing validation protocols that are used to test these devices, the accuracy of available devices is highly variable (34). A small study by Gerin et al found that office blood pressures measured with an automatic device in the absence of a doctor or nurse were more representative of daily ambulatory pressure than physician- or nurse- measurements (35).  Studies have shown that automated pressures  tended to be lower than those recorded by a doctor or nurse. "White-coat hypertension" refers to patients who are diagnosed with hypertension according to clinic pressures measured by medical staff but who have normal ambulatory pressures. "White coat" effects can increase blood pressure by more than 20 m m H g (SBP) and 10 m m H g (DBP) in up to 40% of patients (33). The position of measurement impacts blood pressure readings.  Blood pressures  measured in the supine position show an increase of 0-3 mm H g in SBP and a 2-5 m m H g decrease in D B P compared to those measured in the sitting position (33).  Expected  orthostatic changes upon standing consist of a decrease in SBP (5-10 m m Hg), a rise in D B P (5 m m Hg), and an increase in heart rate by 5-10 beats per minute (33). It is not known i f  11  position of measurement impacts the blood pressure lowering efficacy of antihypertensive agents.  1.1.7.  Dose-response relationships The most fundamental concept in pharmacology and therapeutics is the log-linear  dose-response curve, which quantitatively describes drug actions, whether therapeutic or toxic. In the past, several antihypertensive drugs were commonly prescribed at excessively high doses due to a lack of formal dose-response studies (27).  Dose-response analyses of  drugs are important in order to identify the lowest possible dose that is both efficacious and safe.  Several  factors  can  affect  the  pharmacodynamic  dose-response  curves  of  antihypertensive medications, including population variation and length of therapy (acute versus chronic). Blood pressure decreases can be offset by counterregulatory mechanisms, such as volume retention and reflex tachycardia.  The pharmacokinetic properties of an  antihypertensive drug can also affect the magnitude of its effect depending on the time that blood pressure is measured relative to drug administration. The trough-to-peak ratio is calculated as the blood pressure decrease at the end of the dosage interval divided by the largest blood pressure decrease during the dosage interval, which usually occurs at the time of peak plasma drug concentration.  A s the trough-to-peak ratio gets closer to 1, the  antihypertensive effect is consistently maintained throughout the dosage interval and thus the pharmacologically-generated blood pressure variability is reduced (36). According to the Food and Drug Administration guidelines, a trough effect that is at least 50% of the peak effect is desired (36).  12  1.2 The role of calcium in the cardiovascular system 1.2.1.  Calcium Calcium ions (Ca ) play significant regulatory roles in excitable tissue, from neuronal 2+  transmitter release to muscle contraction. With respect to the cardiovascular system, entry of external calcium through the L-type calcium channels plays a vital role in excitationcontraction coupling in the heart and regulation of diameter in vascular smooth muscle. A resting intracellular calcium concentration of about 150 n M and an extracellular calcium concentration of about 1-2 m M create a large concentration gradient across the cell membrane (37). Thus, calcium ions are sensitive signal transducers.  1.2.2.  C a l c i u m channels i n the cardiovascular s y s t e m  There are two distinct families of calcium channels: intracellular release channels, which allow calcium ions to move into the cytosol from intracellular stores, and plasma membrane calcium channels, which control the entry of calcium ions from the extracellular space. A t the plasma membrane, there are three main types of channels which differ by their regulatory mechanism: voltage-operated channels, store-operated channels and receptoroperated channels (38). The voltage-gated calcium channels are further classified according to their electrophysiological and pharmacological properties (L, N , P / Q , R and T types). In the cardiovascular system, L - and T-type channels predominate. The L-type ("long lasting") channel, which has a large conductance and slow inactivation rate, is the major type of calcium channel in excitable cells and is composed of five subunits. This type is also the main target of clinically available CCBs. In the heart, the function of L-type channels at the sinoatrial and atrioventricular nodes is for pacemaker activity and conduction of impulses, respectively.  In cardiac muscle, action potentials  13  depolarize the plasma membrane which then triggers the opening of L-type calcium channels in the plasma membrane. These localized increases in C a of C a  2+  2+  concentration then lead to release  through ryanodine-receptor channels from the sarcoplasmic reticulum by a process  called "Ca -induced C a 2+  2+  release" (39). The smooth muscle of the vasculature utilizes a  similar process to initiate contraction, except that the release of C a  2+  from intracellular stores  also occurs via inositol trisphosphate (IP ) receptors. The I P receptors permit the alteration 3  3  of force production in the absence of a change in membrane potential because IP acts as a 3  second messenger following binding of agonists to receptors at the plasma membrane ("pharmacomechanical coupling"). Hormones influence C a operated C a  2+  channels.  2+  influx by acting on receptor-  The diameter of the vasculature also depends on several local  factors, including perfusion pressure, p H , and p 0 . 2  T-type calcium channels are found on arterial vascular walls, myocardial conduction tissue (atrioventricular node, sinoatrial node, Purkinje cells) and at neurohormonal release sites (39). While the L-type channels play a major role in myocardial contraction, T-type .channels instead exert their effects on pacemaker function and regulation of vascular smooth muscle tone. T o maintain a low resting level of C a  2+  in the cytosol, C a  2+  is removed via C a 2+  ATPase pumps at the sarcoplasmic reticulum or plasma membrane, and also C a / N a 2 +  +  exchangers.  1.3 Calcium channel blockers 1.3.1.  H i s t o r i c a l aspects  In the 1960s, Fleckenstein et al. discovered the calcium antagonist properties of verapamil and prenylamine in studies using isolated cardiac tissue (40).  These and  subsequent studies elucidated important effects of calcium antagonists on the heart, namely  14  coronary artery dilation and negative inotropic effects that could be reversed by increasing the extracellular C a  concentration.  2+  N o t until the 1970s were the peripheral vasodilator  properties of CCBs discovered. Since then, the array of therapeutic uses of CCBs has grown from angina, hypertension and arrhythmias, to cerebral vasospasm, migraine and Raynaud's phenomenon, among others.  1.3.2.  Classification In theory, calcium antagonists refer to all compounds that potentially inhibit calcium-  dependent processes. Calcium channel blocker drugs belong to a chemically heterogeneous family. According to the W H O classification (41), Group A CCBs consist of agents that have  high affinity  and  specificity  for  the  L-type calcium  channel including  the  phenylalkylamines, cUhydropyridines, benzothiazepines (Table 3). Group B CCBs (Table 4) are less specific L-type channel blockers and they are much less widely used.  15  T a b l e 3: Classification of group A c a l c i u m antagonists Phenylalkylamines Animparnil Desmethoxyverapamil Devapamil Falipamil Gallopamil Rompamil Terodiline Tiapamil Verapamil*  Dihydropyridines Amlodipine Aranidipine Azelnidipine Barnidipine (formerly Mepirodipine) Benidipine Cilnidipine Clevidipine Darodipine Efonidipine Elgodipine Felodipine Flordipine Isradipine Lacidipine Lercanidipine Manidipine Masnidipine Nicardipine • Nifedipine* Niguldipine Niludipine Nimodipine Nisoldipine Nitrendipine Oxodipine Riodipine Silvadipine  Benzothiazepines Clentiazem Diltiazem* Fostedil  1  - prototype agents T a b l e 4: G r o u p B c a l c i u m antagonists Bencyclane Bepridil Caroverine  1.3.3.  Cinnarizine Etafenone Fantofarone  Fendiline Flunarizine Lidoflazine  Lomerizine Perhexiline Prenylamine  Semotiadil TerocUline  Pharmacodynamics of c a l c i u m channel blockers  1.3.3.1. M o l e c u l a r m e c h a n i s m of action Calcium channel blockers  reduce the cytosolic free-calcium  concentrations by  blocking transmembrane calcium influx through L-type calcium channels. calcium channel is composed of five subunits: a,, a , [3, y, and 5. 2  The L-type  Dihydropyridines,  16  benzodiazepines and phenylalkylamines bind to the pore-containing a,-subunit of the L type calcium channel, the former two near the external face of the pore and the latter within the pore.  Ligand-binding sites on the L-type channel have several aflosteric interactions  between each other and with the gating machinery of the channel pore (Figure 1) (42). CCBs are believed to induce a conformation change that influences gating of the in the L type calcium channel (43).  Figure 1. B i n d i n g sites of c a l c i u m channel blocker drugs. (-) denotes a negative allosteric interaction. (+) denotes a positive allosteric interaction. Adapted from Triggle 1991 (42).  17  1.3.3.2. P h y s i o l o g i c a l effects  In general, calcium channel blockers relax arteriolar smooth muscle, resulting in vasodilatation and decreased peripheral resistance. decreases blood pressure.  The decreased afterload  therefore  In addition to peripheral vasodilatation, CCBs increase coronary  blood flow. CCBs also have a negative inotropic effect on the myocardium. Agents that slow the rate of recovery of L-type calcium channels (verapamil, diltiazem) have negative chronotropic and dromotropic effects on the heart's conducting system.  The natriuretic  effect of CCBs may contribute to their ability to lower blood pressure. The most common side effects  associated with CCBs  stem from excessive  vasodilatation (40), and include hypotension, dizziness, headache, flushing and nausea. Constipation (caused by decreased gut motility), peripheral edema (caused by redistribution of extracellular fluid), coughing and wheezing also occur with CCBs. CCB-induced vasodilatation also tends to trigger a reflex increase in sympathetic activity, producing thetachycardia commonly observed with dihydropyridines.  1.3.3.3. P h e n y l a l k y l a m i n e s  Verapamil,  the  prototype  phenyalltylamine,  is  indicated  for  treatment  of  hypertension, angina and arrhythmias (40). O f all the subclasses of C C B s , it has the most cardiac effects. Verapamil's vasodilatory activity is less than that of dihydropyridines, but the negative chronotropic effect of the drug mitigates any reflex tachycardia. The prolongation of refractor)' periods in the cardiac conduction system can lead to heart block or sinus arrest in some patients.  Heart failure precipitated by the negative inotropic effect is another  potentially serious adverse effect associated with verapamil.  18  1.3.3.4.  Benzothiazepines  The benzodiazepine prototype diltiazem is indicated as an antihypertensive and antianginal agent. Diltiazem decreases blood pressure through arterial dilatation. It is less cardiodepressant than verapamil but negative chronotropic effects usually cause a decrease in heart rate (40).  The most  common  side effects  of diltiazem include  headache,  swelling/edema, arrhythmias (first degree heart block), and asthenia (44).  1.3.3.5.  Dihydropyridines  1,4-Dihydropyridines represent the largest subclass of CCBs.  Nifedipine is the  prototype of this group of relatively vascular-selective CCBs. The fall in blood pressure and relief of angina can be attributed to vasodilatation in peripheral arterioles and coronary arteries, respectively (40). The negative inotropic effect of nifedipine is minor at therapeutic doses and is usually overcome  by the  baroreceptor-mediated  sympathetic  reflexes.  Dihydropyridines are associated with exacerbation of myocardial ischemia due to several possible mechanisms: excessive hypotension, coronary vasodilatation of nonischemic areas ("coronary steal"), and increased oxygen demand resulting from increased reflex sympathetic activity. Common side effects of dihydropyridines in the hypertensive population include edema,  headache,  fatigue,  dizziness, constipation, nausea,  flushing,  palpitation  and  tachycardia (44).  1.3.3.6. B e n z i m i d a z o l y l s  While most calcium channel antagonists predominantly block the L-type calcium channel, a newer class of benzimidazolyl tertraline derivatives inhibits both L - and T-type channels (45). One drug from this class is mibefradil, which possesses higher selectivity for T-type than L-type calcium channels. Mibefradil also binds to a unique site on the L-type  19  channel, with negative allosteric interactions with the verapamil, diltiazem and fantofarone binding sites. Mibefradil reduces heart rate but lacks both the negative inotropic effects of other cardio-selective calcium channel blockers and the reflex increases in sympathetic activity associated with some dihydropyridines. Though initially approved in 1997 as a treatment for hypertension and chronic stable angina, mibefradil (Posicor®) was removed from the market in 1998 because of the incidence of serious drug interactions arising from cytochrome P450 inhibition, leading to cardiogenic shock in a small number of patients (46).  1.3.4  Pharmacokinetics of c a l c i u m channel blockers C C B drugs are well absorbed (>90%) from the gastrointestinal tract but undergo  first-pass  hepatic  metabolism resulting in low bioavailability (40).  Drugs in  the  dihydropyridine subclass are highly bound to plasma proteins (> 95%) while diltiazem and verapamil are less so. Early calcium antagonists were all short-acting, with time to maximum concentration occurring within about 2 hours. The rapid decreases in blood pressure gave rise to many side effects, especially tachycardia from reflex sympathetic nervous system activation, flushing, headache and dizziness. Newer CCBs such as amlodipine and slowrelease formulations of older CCBs were developed to produce a more gradual decrease in blood pressure with a longer duration of B P control, increased safety, and less side effects. Extensive metabolism of CCBs occurs via oxidative enzymes in the liver, primarily the 3A4 isozyme of the cytochrome P450 family (47). Half-lives and clearance rates vary considerably among CCBs (Table 5). Dihydropyridines are eliminated in the urine with the majority in the form of inactive metabolites and small amounts in the unchanged form, ft is important to note that pharmacokinetic parameters can vary according to length of administration (acute vs. chronic) and that there is much interindividual variability.  20  T a b l e 5: P h a r m a c o k i n e t i c parameters o f C C B s (44, 47) Drug  Amlodipine Diltiazem immediate release sustained release controlled deliver)' Felodipine extended release Flunarizine Isradipine Lacidipine Nicardipine Nifedipine immediate release prolonged action extended release  Absorption (%)  Bioavailability (%)  PPB (%)  (hours)  >95% 80-90  63 40-67  97.5 70-80  6-10 2-4  >95%  Wellabsorbed >90% ' >90% >90% >90%  7-11  35-50 3.5-6 5-8 (repetitive dosage) 5-7  10-14  5^8  99  2.5-5  11-16  99  2-4  19 days  15-24 14-19 10-17 45-75  95 98 98 95  2-4 3 0.5-1 0.5-2  8.3 7-18 2 2-5  50-70  98  4  15  86  Nilvadipine Nimodipine  >90% >90%  14-19 13  99 99  Nisoldipine Nitrendipine Verapamil immediate release  >90% >90% >90  8 -30 10-20  99 98 >90  Renal Unchanged % 10 2-4  Elim. Met. %  <0.5  70  <1 >1 <1 0.1  65 70 60 -75  60  6  10  2.15 1-2 (early) 8-9 (terminal) 1-2 1-2 1-2  7-18 2  <1 0  -75 15  10-12 6-15 2.8-7.4 4.5-12 (repetitive dosage) 5.3-9.6  <1 0.1 3  -75 45 86  sustained 20-35 4-8 release P P B = percentage protein binding Renal unchanged % = % excreted renally as unchanged drug E l i m Met % = % eliminated as metabolites tmax = time to maximum plasma concentration t i / 2 = half life  1.3.5  tl/2 (hours)  C l i n i c a l use of c a l c i u m c h a n n e l blockers Calcium channel blockers are used widely i n clinical practice to lower elevated blood  pressures and manage angina and arrhythmias.  Although C C B s are generally considered  21  second- or tliird-line therapies in the treatment of hypertension in subjects under 60 years of age, CCBs have been recommended as first-line treatment for treatment of isolated systolic hypertension in the elderly (10). In 2003, the C C B amlodipine (Norvasc®) was the fourth top selling prescription drug in the world (48). Nevertheless, controversy surrounds the use of C C B s in the treatment of chronic hypertension,  fn 1995, two retrospective studies showed increased risks of myocardial  infarction and mortality in those taking calcium channel blockers (49, 50).  Psaty's case-  control study showed that among users of CCBs (with or without a diuretic) the risk ratio of myocardial infarction was 1.62 (95% CI 1.11-2.34) compared with those taking a diuretic alone, and  1.57 (95% CI 1.21-2.04) compared with users of beta-blockers (49). Pahor et  al's cohort study of elderly hypertensive patients showed that in comparison with betablockers, the relative risks for mortality associated with use of verapamil, diltiazem, and short-acting nifedipine were 0.8 (95% Cf 0.4-1.4), 1.3 (95% CI 0.8-2.1), and 1.7 (95% CI 1.12.7), respectively (50). A meta-analysis by Furberg et al. showed that the use of short-acting nifedipine was associated with increased mortality in patients with coronary heart disease (overall R R 1.16, 95% CI 1.01-1.33) (51). A dose-response relationship was present, with the risk rising sharply with doses above 80 mg/day of nifedipine (RR at 80 mg/day was 2.83; 95% Cf 1.35-5.93).  The potential mechanisms by which calcium channel blockers may  increase mortality are varied, including pro-ischemic effects, marked hypotension, negative inotropic effects, prohemorrhagic effects and activation of the sympathetic- and reninangiotensin systems, thereby predisposing to arrhythmias (51). T o assess whether similar effects were observed with long-acting CCBs, another case-control study was performed (52).  The study showed no difference in risk of  cardiovascular events between beta-blockers and long-acting CCBs, but there was a marked  22  increase in risk in patients who took short-acting CCBs compared with beta-blockers (adjusted odds ratio 3.88, 95% C I 1.15-13.11) and long-acting CCBs (risk ratio 8.56, 95% C I 1.88-38.97).  However, it is important to note that all observational studies are subject to  confounding since the patients may differ fundamentally between treatment groups. For example, in the aforementioned studies, the prevalence of coronary heart disease was greater in the groups taking short-acting CCBs (53). Links between use of CCBs and increased risk of cancer, major hemorrhage and suicide had been suggested initially from observational studies, although these associations were contradicted in subsequent studies (53). The debate over calcium channel antagonist safety was addressed further with respect to the potential conflict of interest arising from authors' relationships with the pharmaceutical industry (54). This study of calcium-channelantagonist articles published between March 1995 and September 1996 found that 96% of the authors who were supportive of the use of calcium channel blockers had financial relationships with C C B manufacturers, compared to 60% of neutral authors and 37% of critical authors (p<0.001). Numerous large, long-term clinical trials of CCBs have been carried out (Table 6) but the debate over the efficacy and safety of CCBs still rages. The findings of Pahor et al.'s systematic review demonstrating that CCBs given as  first-line  antihypertensives are  associated with a higher incidence of major cardiovascular events compared with other drug classes (55) have been tempered by the results of recent clinical trials ( A S C O T [56], F E V E R [57]) that show CCBs in combination with other antihypertensives may have benefits in terms of cardiovascular events and mortality.  23  Table 6: L o n g term clinical trials of C C B s for treatment of hypertension Study  Trial Design & Population  Interventions and M a i n Result  ABCD 1998 (58)  R/DB N=950 patients with type 2 diabetes mellitus; [N=450 had D B P >90 mm Elg (mean age: 57.5 y) and N=480 were normotensive]  fntensive (target of D B P 80-89 m m Hg) vs. moderate therapy (target of D B P 75 mm Hg) and nisoldipine vs. enalapril as l -Une antihypertensive treatment, with possible addition of metoprolol and hydrochlorothiazide. Mean follow-up: 5 years. In the hypertensive group, the incidence of fatal and nonfatal myocardial infarctions was significantly higher among those receiving nisoldipine compared with those receiving enalapril (RR 5.5; 95% CI 2.1-14.6; p < 0.001).  ALLHAT 2002 (59)  ASCOTBPLA 2005 (56)  CONVINCE 2003 (60)  st  R/DB Chlorthalidone vs. lisinopril vs. amlodipine (doxazosin N=33 357 (mean arm terminated), with possible addition of atenolol, age 67 y), SBP > clonidine, or reserpine. Mean follow-up: 4.9 years. 140 mm H g N o statistically significant difference in total C H D and/or D B P >90 events or total mortality between the different mm H g plus one treatments, fncreased incidence of heart failure with additional risk amlodipine compared to chorlathlidone treatment (RR factor for C H D 1.38; 95% CI 1.25-1.52, p<0.001) R/open-label; Amlodipine (plus perindopril as required) vs. atenolol N=19 257 (mean (plus bendro flume thiazide as required). Mean followage 63 y), SBP > up: 5.5 years. N o statistically significant difference in 160 mm H g non-fatal M f + fatal C H D (primary endpoint). and/or D B P > Amlodipine-based regimen showed reduction in total 100 m m H g i f cardiovascular events and procedures (Hazard ratio untreated, or SBP [HazR] 0.84, 95% CI 0.78-0.90), all-cause mortality > 140 m m H g (HazR 0.89, 95% Cf 0.81-0.99), fatal and non-fatal and/or D B P > stroke (HazR 0.77, 95% CI 0.66-0.89). 90 mm H g i f treated; plus at least 3 other C V risk factors R/DB Hydrochlorothiazide or atenolol vs. verapamil C O E R , N=16 602 (mean with possible addition of other study drugs to achieve age 66 y); B P conrol. Mean follow-up 3 years. N o statistically patients had > 1 significant difference in C V death, M I , stroke between Verapamil had higher incidence of heart C V risk factor in groups. addition to failure (HazR 1.30, 95% CI 1.0-1.69, p=0.05) and death or hospitahzation for bleeding unrelated to hypertension stroke (HazR 1.54, 95% CI 1.15-2.04, p=0.003). Study terminated early by sponsor for commercial reasons.  24  Study FACET 1998 (61)  FEVER 2005 (57)  HOT 1998 (62)  INSIGHT 2000 (63)  INVEST 2003 (64)  Trial Design & Population R/open-label; N=380 (mean age: ?y); patients with type 2 diabetes mellitus and SBP >140 m m H g or D B P 90 mm H g  R/DB N=9800 Chinese patients (mean age 61.5 y), SBP 160-210 m m Fig, D B P 95-115 m m Hg, plus one additional CV risk factor R/DB N=18 790 (mean age: 61.5 y), D B P 100-115 m m H g  R/DB N=6321 (mean age: 65 y); SBP > 150/95 m m H g or SBP > 160mm Hg, + >1 additional CV risk factor R/open-label, masked endpoint; N=22 576, (mean age 66 y); patients had C A D and hypertension  Interventions a n d M a i n Result Fosinopril vs. amlodipine, with possible addition of other study drug to achieve B P control. Mean followup 2.8 years. N o statistically significant differences between the two groups in serum lipids, glucose control & renal function parameters. Amlodipine group had a significantiy greater reduction in SBP than fosinopril group. Fosinopril group had significandy lower rate of the combined endpoint of stroke, acute myocardial infarction, or hospitalization for angina (RR 0.49; 95 CI 0.26-0.95; p = 0.030) compared to amlodipine. Felodipine + H C T Z vs. placebo + H C T Z , with possible addition of other study drugs. Mean followup 40 months. Felodipine group had reduction in nonfatal stroke (HazR 0.74, 95% CI 0.59-0.91), total C V events (HazR 0.73, 95% CI 0.51-0.86), all-cause death (HazR 0.69, 95% CI 0.54-0.89).  Randomized to 3 D B P targets: <90 m m H g vs. <85 m m H g vs. <80 m m Fig (using felodipine 5 mg as the initial drug in a stepped therapy), and also randomized to acetylsalicylic acid vs. placebo. Mean follow-up: 3.8 years. N o statistically significant difference in cardiovascular events between B P target groups, except in subgroup of patients with diabetes mellitus, which showed lower risk of events in <80 m m H g group vs. <90 m m Fig group. Nifedipine GITS vs. hydrochlorothiazide + amiloride, with possible addition of enalapril or atenolol. Mean follow-up: 3.1 years. N o statistically significant difference in combined primary endpoint (cardiovascular death, myocardial infarction, heart failure, or stroke; R R 1.10 [95% CI 0.91-1.34]; p=0.35)  Verapamil SR (plus trandolapril as required) vs. atenolol (plus H C T Z as required). Mean follow-up: 2.7 years. N o statistically significant difference between treatment strategies in primary outcome (composite of death, nonfatal M I , nonfatal stroke)  25  Study MIDAS 1996 (65)  NICS-EH 1999 (66)  NORDIL 2000 (67)  STOP-2  1999 (68)  SYST-CHINA 1998 (69)  Trial Design & Population R/DB N=883 (mean age: 58.5 y), D B P 90-115 m m H g  Interventions and M a i n Result Isradipine vs. hydrochlorothiazide, with possible addition of open-label enalapril. N o significant differences between the two groups in isradipine group in progression of carotid mtimal medial thickness. Isradipine group had a non-statistically significant increase in major cardiovascular events compared to diuretic group (RR 1.78; 95% CI 0.943.38; p = 0.07).  R/DB N=429 (mean age: 69.8 y), SBP 160-220 m m H g and D B P <115 mm H g R/open-label, masked endpoint; N = 10,881 (mean age: 60 y); D B P > 100 mm H g  Nicardipine SR vs. trichlormethiazide. Follow-up: 5 years. N o statistically significant difference between the two groups in rate of cardiovascular events (p=0.923).  R/open-label, masked-endpoint; N=6614 (mean age: 76 y), SBP >180 m m H g , D B P >105 m m Hg, or both  Conventional drugs (atenolol, metoprolol, pindolol, or hydrochlorothiazide plus amiloride) vs. newer drugs (either A C E inhibitors enalapril or lisinopril, or CCBs felodipine or isradipine). Follow-up: 5 years. N o statistically significant difference in primary combined endpoint (fatal stroke, fatal myocardial infarction, and other fatal cardiovascular disease; p=0.89).  DB/PC N=2394 (mean age: 67 y), SBP 160-219 m m H g and D B P <95 mm H g  Assigned by alternation to active treatment (using nitrendipine as initial drug in stepped therapy with add-on of captopril or hydrochlorothiazide or both), vs. placebo. Median follow-up: 3.0 years. Statistically significant reduction in all cardiovascular endpoints combined (fatal + non-fatal) in active treatment group compared to placebo but no difference in cardiac endpoints.  Diltiazem vs. beta blockers, diuretics or both. Mean follow-up: 4.5 years. N o statistically significant difference in combined primary endpoint (fatal and non-fatal stroke, fatal and non-fatal myocardial infarction, and other cardiovascular death; p=0.97) between groups.  26  Study SYST-EUR 1997 (70)  Trial Design & Population R/DB/PC N=4695 (mean age: 70.3 y), sitting SBP 160219 m m H g and sitting D B P <95 mm H g  TOMHS 1996 (71)  R/DB/PC N=902 (mean age 54.8 y), D B P 90-99 mm H g or 85-99 m m H g after withdrawal in those taking 1 antihypertensive drug  VALUE 2004 (72)  R/DB N=15 245 (mean age 67 y); SBP 160-210 m m H g and D B P < 115 mm H g ; patients at high C V risk  VHAS 1997 (73)  Interventions a n d M a i n Result Active treatment (using nitrendipine as initial drug in stepped therapy with add-on of enalapril or hydrochlorothiazide or both) vs. placebo. Median follow-up: 2 years. Statistically significant reduction in fatal + non-fatal stroke (p=0.003) and total cardiovascular endpoints (p<0.001) in active treatment group compared to placebo. N o difference in allcause and all-cardiovascular mortality. Chlorthalidone vs. acebutolol vs. doxazosin vs. amlodipine vs. enalapril vs. placebo (nutritional/hygienic intervention administered to all patients). Mean follow-up: 4.4 years. N o significant difference between all drug treatments aggregated and placebo in # of patients experiencing at least 1 event (coronary heart disease and cardiovascular disease events).  Valsartan vs. amlodipine (plus addition of other antihypertensives i f required). Mean follow-up: 4.2 years. Primary composite endpoint of cardiac morbidity and mortality + all-cause mortality showed no statistically significant difference between regimens. Valsartan regimen had higher incidence of M I compared with amlodipine (HazR 1.19, 95% CI 1.021.38,p=0.02) r R/DB (1 6 Verapamil vs. chlorthalidone, with possible addition of months, then captopril or switch to free therapy in non-responders. open-label); A t 2 years follow-up, there were no statistically N=1414 (mean significant differences in # of patients with D B P age: 53.2 y), SBP normalization and in cardiovascular events between > 160 m m H g the two groups. and D B P 95-114 m m Tig st  1.4  Systematic reviews  1.4.1  What are systematic reviews? Systematic reviews answer specific clinical questions in depth, by synthesizing the  results of primary research and using explicit methods that limit bias (74). Developing systematic reviews has been considered a "fundamental scientific activity." (75)  While  27  researchers can use reviews to identify new hypotheses and to refine their approaches, health care providers and policy makers can apply the results of such reviews towards rational decision making (75). However, it is important to note that research evidence is only one of many components that are involved in decision making in each patient-clinician encounter (74). A well-formulated question specifies the following five variables: 1) types of study designs, 2) types of participants, 3) types of interventions, 4) types of outcomes, and 5) types of control comparisons. For example, the question addressed in this review can be stated as: In randomized controlled trials, how much do calcium channel blockers lower blood pressure compared with placebo in patients with primary hypertension? Systematic reviews can be categorized as qualitative or quantitative.  In the latter  case, the review is referred to as a meta-analysis and involves quantitatively combining results from individual studies. Using statistical methods to combine the data increases the power and precision of the estimate of the effect size. The other advantages of a systematic review are well established.  Health care  workers, researchers and consumers are often deluged with unmanageable amounts of information  (76).  Systematic reviews concisely and efficiently integrate  up-to-date  information such that important effects of health care can be identified promptly (74). In addition, systematic reviews can establish the degree of generalizabiHty and consistency of the findings across populations and different treatments (76).  Lastly, since systematic  reviews explicitly apply scientific principles that are designed to reduce random and systematic errors, the results are more likely to be reliable and accurate than individual studies or narrative reviews.  28  Like all secondary research, systematic reviews are retrospective studies and are hence subject to bias. Meta-analyses can produce inflated effect sizes due to publication bias, which describes a systematic error arising from the tendency of positive-result trials to be more likely to be published. (and often numerous times) than negative-result trials. Systematic reviews are designed to reduce bias by 1) including studies with multiple publications only once; 2) using explicit, reproducible criteria for selecting studies and for extracting data; and 3) having more than one independent reviewer to perform these tasks (76).  1.4.2  How does a systematic review differ from a narrative review? "Narrative" or "traditional" reviews of research have always been common in  medical literature. Frequently, these reviews have a broad scope and are written by experts in the field who tend to focus only on studies that support their own view. A s well, the majority of narrative reviews lack explicit, systematic methods with respect to how studies are selected and integrated. For these reasons, there is great likelihood of introducing bias. Since the methodologic quality of medical review articles is quite variable, there are concerns about the validity of their conclusions (77). Conversely, a systematic review utilizes an explicit search strategy that is applied to several sources. Systematic reviews are focused on a specific clinical question and studies are selected using pre-specified inclusion criteria. This approach allows a more objective view than traditional narrative reviews. Thus, a major advantage of a systematic review is that the same rigor as that expected from primary research is applied. Consequently, carrying out a systematic review requires much more time and effort than a narrative review.  29  1.4.3  T h e Cochrane Collaboration The Cochrane Collaboration (CC) is an international organization that aims to  prepare, maintain and make accessible systematic reviews of the effects of health care. Established in 1992, the organization was named after Archie Cochrane, a British epidemiologist who recognized the importance of having reliable reviews of evidence to make informed decisions about heath care. The C C espouses the following ten principles: collaboration, building on the enthusiasm of individuals, avoiding duplication, minimizing bias, keeping up to date, ensuring relevance, ensuring access, continually improving the quality of its work, continuity and enabling wide participation (78). The systematic reviews produced within the C C are published electronically in the Cochrane Library, which is updated quarterly.  The format of a Cochrane review allows readers to find the results  quickly and allows them to submit comments and criticisms. A s well, the format facilitates updating reviews with new or previously missing trials. A study that compared Cochrane reviews with articles published in paper-based journals revealed that Cochrane reviews tended to have greater methodological rigor and were more likely to be updated (79). 1.4.4  A i m of this systematic review A t the present time, the blood pressure-lowering efficacy of the different subclasses  of calcium channel blockers and the individual drugs within each subclass is not known, and a systematic review of the short-term dose-related blood pressure lowering effect of calcium channel blocker drugs has not been previously performed.  The aims of this systematic  review are to determine these dose-related effects in patients with primary hypertension and to establish dose equivalencies of different drugs within the calcium channel blocker family. The information derived from this review should facilitate future reviews of head-to-head  30  comparisons with other drug classes and assist clinicians in choosing optimal doses calcium channel blockers.  2.  PROTOCOL The protocol for this systematic review was finalized in March 2000 and first  published in Issue 2, 2002 of the Cochrane Library (80) to outline the scientific methods that would be employed. The methodology was based on the Cochrane Reviewers' Handbook (76) and on a previous systematic review that assessed the blood pressure lowering efficacy of thiazide and loop diuretics (81).  2.1 Objectives Primary objective: T o quantify the dose-related effects of various doses and types of calcium channel blockers on systolic and diastolic blood pressure versus placebo in patients with primary hypertension.  Secondary objectives: 1. T o determine the effects of calcium channel blockers on variability of blood pressure. 2. T o determine the effects of calcium channel blockers on pulse pressure. 3. T o quantify the dose-related effects of calcium channel blockers on heart rate. 4. T o quantify the effects of calcium channel blockers in different doses on withdrawals due to adverse events.  2.2 Methodology 2.2.1  Types of studies Included studies must be randomized controlled trials (RCTs) and their design must  meet the following criteria:  32  •  random allocation to calcium channel blocker monotherapy group (s) and a parallel placebo control group  •  double-blind  •  duration of follow-up of at least three weeks  •  office blood pressure measurements were made at baseline (following washout) and at one or more time points between 3 to 12 weeks after starting treatment  2.2.1.1.  Why are only randomized controlled trials included?  Randomized controlled trials are prospective, quantitative, comparative experiments in which patients are allocated randomly to receive one of two or more interventions and the ensuing clinical outcomes are compared. Since all patients have the same chance of being assigned to each group, any baseline differences that arise between groups occur by chance. Random allocation reduces the risk of imbalance of both known and unknown baseline factors.  In this way, the investigators can isolate the impact of the intervention being  studied, thereby rrunimizing the influence of other factors that could affect the outcome of the study. Proper randomization can be achieved in many ways; usually random number tables or computers are used to produce a pre-specified randomization list. Randomization controls for selection bias (that is, systematic differences in the way comparison groups are assembled), although efforts to randomize can be undermined i f allocation is not concealed.  2.2.1.2.  Why is a parallel placebo arm necessary?  Included RCTs must have a parallel placebo arm as the control comparison since a significant but variable placebo effect (in the range of 0 to 10 mm H g decrease in SBP and 0 to 8 mm H g in D B P ) exists for antihypertensive drugs (82). Since the objective of this  33  review is to assess the dose-related blood pressure lowering efficacy of calcium channel blockers compared to placebo, the overall effect of calcium antagonist treatment is obtained by subtracting the effect in the placebo group from the effect in the active treatment group. Having a placebo group not only controls for bias, but also minimizes the effect of "regression to the mean", a statistical phenomenon that describes the spontaneous withinpatient variation and/or variation due to measurement errors.  This effect has been  demonstrated in hypercholesterolemic and hypertensive patients by a shift from high to low values with time, independent of treatment (83). Thus, regression to the mean is responsible for many false positive results in experiments that involve before-after measurements of a continuous variable (83). For the purposes of this review, cross-over trials are only accepted if data are reported for the first 3-12 week phase of treatment with active drug versus a placebo arm. Thus, this is effectively the same as a parallel-design trial.  2.2.1.3.  Why is blinding (masking) necessary?  Ascertainment bias involves systematic distortion of the results of the trial by knowledge of group assignment, the source of which can be the observers, participants, or assessors (84). Bknding, also known as masking, reduces the risk of this form of bias, fn a double-blind R C T , both the participants and the investigators do not know the group assignment of each participant.  Proper blinding is achieved by testing an intervention  against a matched placebo, with identical taste, smell, and appearance as the active drug. Methods for assessing the success of blinding are available.  However, the success of  blinding in hypertension trials has rarely been tested.  34  2.2.1.4  Why is a baseUne measurement subsequent to or during a washout/placebo run-in period important?  The main purpose of a washout period is to eliminate carry-over effects from previous drug administration prior to commencing the trial. Commonly, the washout period is comprised of or is followed by a single-blind placebo "run-in" before randomization. The most accurate estimates of the baseline blood pressure and heart rates are obtained at the end of this period. This pre-study period also aids in the selection of participants who are truly hypertensive, as those who do not meet the inclusion criteria cut-off after the washout/run-in do not continue in the study. Repetitive measurements before study entry allow blood pressures to stabilize, and thus, the problem of regression to the mean is mirLimized.  2.1.1.5.  Why is the 3-12 week window selected?  Based on a previous systematic review (81), a follow-up of 3-12 weeks during treatment represents an appropriate window of time from which to extract outcome data. A rninimum of three weeks is usually necessary for the effect of therapy to become maximal and stable. For practical reasons, the upper duration limit of 12 weeks was chosen since many antihypertensive drug trials involve dose titration or addition of other drugs in those who fail to achieve a target blood pressure after a certain length of time on treatment. A s well, a shorter duration allows the data to include the maximum number of patients, as withdrawals and drop-outs, which can confound results, occur at higher rates in trials with long duration.  2.2.2.  T y p e s of participants Participants must have had a baseline systolic blood pressure > 140 m m H g and/or a  diastolic blood pressure > 90 m m Fig, measured in a standard way. Patients with creatinine  35  levels greater than 1.5 times the normal level were excluded. This criterion thereby excludes patients with secondary hypertension due to renal failure.  Patients who were pregnant  during the study were excluded since pre-eclampsia is considered a form of secondary hypertension. Participants who took medications that affect blood pressure other than the study medications were excluded. Participants were not restricted by age, gender, baseline risk, or any other co-morbid conditions.  2.2.3.  Types of interventions The intervention of interest is monotherapy with any calcium channel blocker, as  listed in Table 4 and 5. Trials in which titration to a higher dose is based on blood pressure response were not eligible i f the titration occurred before three weeks of treatment. This restriction is in place because dose-response  relationships cannot be analyzed i f patients witiiin each  randomized group are taking different doses.  However, trials in which a response-  dependent titration took place during or after the 3-f2 week interval were eligible i f pretitration data were given. In cases of forced (i.e., response-independent) titration, data from the highest dose given for the specified duration of time, provided this duration was > 3 weeks, were extracted.  2.2.4.  Types of outcome measures  Primary: Change from baseline in trough and/or peak systolic and diastolic blood pressure at 3-12 weeks treatment, compared to placebo. If blood pressure measurements were available at more than one time within the acceptable window, the weighted means of blood pressures taken in the 3-12 week range were used.  36  Secondary: 1.  Standard deviation of the change in blood pressure compared with placebo.  2.  Change in standard deviation of blood pressure compared with placebo.  3.  Change in pulse pressure compared with placebo.  4.  Change in heart rate compared with placebo.  5. Number of patient withdrawals due to adverse events compared with placebo.  2.2.5.  Search strategy for identification of studies T o identify randomized placebo-controlled trials of calcium channel blockers,  Medline (Jan. 1966-June 2003), E M B A S E (fan. 1988-June 2003), C I N A H L (Jan. 1982-June 2003), the Cochrane Controlled Clinical Trials Register, and bibliographic citations were searched. Previously published meta-analyses on dose-response o f calcium channel blockers were used to help identify references to trials. N o language restrictions were applied. The following search strategy was utilized to identify randomized, placebo-controlled trials of calcium channel blockers for hypertension.  This approach was a modified,  expanded version of the standard search strategy of the Hypertension Review Group, with additional terms related to calcium channel blockers in general (85). A " / " at the end of a term signifies a Medical Subject Heading (MeSH) term; "exp" means the term is exploded (all M e S H terms nested under the exploded term are included); "pt" denotes publication type; "mp" following a term will retrieve citations that contain the term in the tide, abstract or key words; " o r / " followed by a number range will retrieve the group of citations that fall under any of the corresponding terms.  The symbol "$"  represents a wildcard character used to search for multiple forms of a term. The search  37  modifier "adj" plus a number between two terms returns records that contain the two terms within the specified number of words of each other. 1. randomized controlled trial.pt 2.  randomized controlled trial$.mp  3.  controlled clinical trial.pt  4.  controlled clinical trial$.mp  5. random allocation.mp 6.  exp double-blind method/  7.  double-blind.mp  8.  exp single-blind method/  9.  single-blind.mp  10. or/1-9 11. exp animal/ 12. 10 not 11  13. clinicaltrial.pt 14. clinical trialf.mp 15. exp clinical trials/ 16. (clinf ad)25 trtal$).mp 17. ((singl$ or doubl$ or tripl$ or trebl$) adj25 (blind$ or mask$)).mp 18. randorn$.mp 19. exp research design/ 20. research design.mp 21. or/13-20  38  22. 21 not 11 23. 22 not 12  24. comparative stud$.mp 25. exp evaluation studies/ 26. evaluation studfi.mp 27. follow-up stud$.mp 28. prospective studf.mp 29. (control! or prospective or volunteer!). 30. or/24-29 31. 30 not 11 32. 31 not (12 or 23) 33. 12 or 23 or 32  34. exp calcium channel blockers/ 35. calcium channel blocker$.mp 36. calcium channel antagonist$.mp 37. calcium antagonist!.mp 38. amlodipine.mp 39. anipamil.mp 40. aranidipine.mp 41. azelnidipine.mp 42. barnidipine.mp 43. bencyclane.mp  44. benidipine.mp 45. bepridil.mp 46. cilnickpine.mp 47. cinnarizine.mp 48. clentiazem.mp 49. clevidipine.mp 50. darodipine.mp 51. desmethoxyverapamil.mp 52. devapamil.mp 53. diltiazem.mp 54. dopropidil.mp 55. efonidipine.mp 56. elgodipine.mp 57. etafenone.mp 58. falipamil.mp 59. fanofarone.mp 60. felodipine.mp 61. fendiline.mp 62. flunarizine.mp 63. fostedil.mp 64. gallopamil.mp 65. isradipine.mp 66. lacidipine.mp 67. lercanidipine.mp  68. Hdoflazine.mp 69. lomerizine.mp 70. manidipine.mp 71. masnidipine.mp 72. mibefradil.mp 73. nicardipine.mp 74. nifedipine.mp 75. niguldipine.mp 76. nilvadipine.mp 77. niiriodipine.mp 78. nisoldipine.mp 79. nitrendipine.mp 80. perhexiline.mp 81. pranidipine.mp 82. prenylamine.mp 83. riodipine.mp 84. ronipamil.mp 85. semotiadil.mp 86. silvadipine.mp 87. terodiline.mp 88. dapamil.mp 89. verapamil.mp 90. or/34-89  91. exp hypertension/ 92. hypertension.mp 93. exp blood pressure/ 94. blood presure.mp 95. or/91-94  96. 90 and 95 97. 33 and 96 98. placebo$.mp 99. 97 and 98  2.2.6.  Study Selection The initial search of all the databases was performed to identify citations with  potential relevance.  A l l citations were stored and categorized in the Reference Manager  program. The initial screen of these abstracts excluded articles whose tides and/or abstracts were clearly irrelevant.  The full texts of remaining articles were then retrieved (and  translated into English where required). The bibliographies of pertinent articles, reviews and texts were searched for additional citations.  T w o independent reviewers assessed  the  eligibility of the trials using a trial selection form (Appendix A ) . A third reviewer resolved discrepancies. Trials with more than one publication were counted only once.  2.2.7.  D a t a Extraction Data were extracted independendy  by two reviewers using a standard  form  (Appendix B), and then cross-checked. If data were presented numerically (in tables or text) and graphically (in figures), the  numeric data were preferred  because of possible  42  measurement error when estimating from graphs. A l l numeric calculations and extractions from graphs or figures were confirmed by a second reviewer. The position of the patient during blood pressure measurement may affect the blood pressure lowering effect. However, in order not to lose valuable data, i f only one position was reported, data from that position were extracted. When blood pressure measurement data were available in more than one position, sitting blood pressure was the first preference. If only standing and supine blood pressures are available, that of the standing position was used. In case of missing information in the included studies, investigators were contacted (using e-mail, letter and/or fax) to obtain the missing information. In the case of missing standard deviation of the change in blood pressure or heart rate, the standard deviation was imputed based on the information in the same trial or from other trials using the same drug and dose. The following hierarchy (listed from high to low preference) was used to impute standard deviation values: 1.  pooled standard deviation calculated either from the t-statistic corresponding to an exact p-value reported (86) or from the 95% confidence interval of the mean difference between treatment group and placebo  2.  standard deviation of change in blood pressure/heart rate from a different position than that of the blood pressure/heart rate data used  3.  standard deviation of blood pressure/heart rate at the end of treatment  4.  standard deviation of blood pressure/heart rate at the end of treatment measured from a different position than that of the blood pressure/heart rate data used  5.  standard deviation of blood pressure/heart rate at baseline (except i f this measure is used for entry criteria)  43  6.  weighted mean standard deviation of change in blood pressure/heart rate calculated from at least 3 other trials using the same drug and dose  7. weighted mean standard deviation of change in blood pressure/heart rate calculated from other trials using the same drug 8. weighted mean standard deviation of change in blood pressure/heart rate calculated from all other trials (any drug and dose)  2.2.8.  Quality Assessment Assessing the quality of trials in a systematic review is important in order to  determine the validity of its findings. The quality of reports of randomized trials can be incorporated into a meta-analysis in a number of ways.  In a sensitivity analysis, studies  below a certain quality threshold are removed to determine the effect on the overall estimate. This approach can be used to attempt to explain heterogeneity between trial results. Studies can also be weighted according to quality; with this approach, studies of lower quality have less influence on the overall estimate (87). One of the difficulties with assessing trial quality is that we must rely upon information provided in the written report. T o address this issue, the Consolidation of the Standards of Reporting Trials ( C O N S O R T ) statement issued by a group of researchers and journal editors has aimed to improve the quality of reporting of randomized controlled trials (87). Two quality assessment scales are utilized commonly in systematic reviews: the Cochrane Collaboration approach and the Jadad method. The quality of all included trials was assessed by two independent reviewers using these two approaches.  44  2.2.8.1.  T h e Cochrane approach for assessment of allocation concealment When studies are entered into the RevMan program, the Cochrane quality  assessment scale based on allocation concealment is a default field for the "included trial characteristics" table (76). Each trial in the systematic review is assigned a grade (A, B , C, or D): Grade A : Adequate concealment Adequate concealment can be executed by the following methods: centralized (central office unaware of subject characteristics) or pharmacy-controlled randomization; pre-numbered or coded identical containers that are administered serially to patients; on-site computer system with allocations kept in a locked computer file that can be accessed only after patients enter; sequentially numbered, sealed, opaque envelopes. Grade B : Uncertain Grade B is assigned to trials in which the allocation concealment is not reported, or despite a description that reports adequate concealment (the use of a list, table or sealed envelopes), there are other features that lead the reviewer to be suspicious. Grade C : Clearly inadequate concealment Inadequte allocation concealment consists of the following methods: alternation; use of case record numbers, dates of birth or date at which the patient is invited to participate in the study; any procedure that is transparent before allocation, such as an open list of random numbers.  While these methods in theory produce random groups, knowledge of group  assignment before enrollment can affect the decision to enroll or not enroll the patient. Grade D : N o t used Allocation concealment was not used to assess validity.  45  2.2.8.2.  Jadad M e t h o d Jadad developed a validated tool to assess trial quality (88). This tool is easy to use  and gives consistent measurements. This five-point scoring system is outlined as follows:  1. Was the study described as randomised? (Yes=+1; No=0) If "Yes", was the method of randomization well described and appropriate? (Yes = + 1, N o = -1, Method not reported = 0) 2. Was the study described as double-blind? (Yes= + 1; No=0) If "Yes", was the method of double blmding well described and appropriate? (Yes = + 1, N o = -1, Method not reported = 0) 3. Was there a description of withdrawals and dropouts in each group and the reasons? (Yes= + l ; N o = 0 ) A score of 0-2 reflects low quality, while a score of 3-4 indicates moderate quality and a score of 5 represents a high quality study. 2.2.9.  D a t a Analysis Data synthesis and analyses were done using the Cochrane Collaboration software,  Review Manager (RevMan) 4.2.8. This program allows reviewers to construct systematic reviews according to the Cochrane publication format. RevMan also allows the data to be displayed in a "forest plot".  46  Reve iw: Bo lod pressure o lwern ig eficacy of cac lu im channel bo lckers for "pm ri ary hypertenso in Comparsion.: -01D ' ose Amo ldp in ie :ys. pa lcebo Outcome: ;01 Change n i Systolic Bo lod Pressure Study. . Amo ldp in ie. Pa lcebo VWD (fixed) Weg iht VVtyD i (fixed) on subc -ategory .,N ' hMeanS ;iD) N-: Mean (SD). ' • 95% CI '. .% 95% CI ' Year- Order 01 12 .5 rrg iVca l y; '44 - 4 . Frcik1 .988 ^ 48 -5 60 di70) *45 -6 10(14 10). .'6 0 . 7 8 , 5 . 7 8 ) " 1 9 8 8 0 4 Mehta9 l93 -7 80(1440) 39 -3 70(13 60)?2. -4 o • -10 28, Z. 081 "1993 0 Subtotal 9 (5% CI)' • - 88" .11-16 -1 i44,' -S: 46~';2.5 ' 7] " 8 4 Test for heterogenetiy: = Chi12 . 3", 02 . 7)F" = 187 .% Test for overal effect:07 .Z 1= (P0= 4 .8) 02 25 mg/day• Trick 198846' -8 7DJ1170) 45 -6 10(14 10) 633. -2 say -'?- 93,' 2. 73] 1988 M ' ehta^1993 40 -IO -7 30 (1440) 39 -3 70(13 60) 472 -3 60 -9. ' 78, 2'. 58]' -1"? 'o 0 '•69 "Fs riKman 1995„ 72 70(1500) 00(18 70) 57772 -9 70 -is 31 -4.09] 1995 0 -i • Subtotal 9 (5% CI) 158 1S3 16 -5 30 -8.58, -2.03)' % Test tor heterogenetiy: = Chi36 . 4,= 2 (P 01 .6 = )r = 45. Test lor overal effect3 :Z '1 .7=(P00 = .02) 03 5 mg/day Frc ik 1988 48 -18 70(1040) 45 -6 10(14-id) 10) •702: 12 60 -'17 66 -7.54]' 1 '988 ,1 « Mroczek1988. 10 -16 70 (i 7 ) 5 450 (2 a 050 -21 20 -40 2S -2.IS]' 1988 2 10 Metra i1 '993 44 -8 60 (103 0) 39 -3 70(13 60)-6 55 -4 90 -10 14 03 - 41 1993 3 Kuschnri. 1996 77 -17 SO (1200) 76 -z 10(14 10)10) 1 . 0 44. -IS 40 -19 5S -11.-25] 1996 4 -* Farsarig' 2001 8 ; 4 -is Z0 (1350) 83 -0 90(14 1 0 ? 6 . 1 4 3 0 1 8 4 9 . -rlO.il] 2 001 5 # Poor2001 110 -10 80 (1040) -3 00(14 ib) 1727 -7 80 -11 03 -4.57] 2 . 001 i6^ -iis ' Cnrysant 2003 183 -10 3011 (040) 6S -0 80(14 10) 1 2 8 3 -9 1 3 2 4 -5.76]. . 2 003 s o -9.OS] .t Subtotal 9 (5% CI) 556 428 6486 -10 72 -12 38 Test for heterogenetiy: = C1 h .* 76 .7df - 6= (P00 . 07), I60 .= %6 Test for overale ' ffect: 1Z 26 .1=(P <"0 0 .0001) 04 10 mg/day «• ; Lciata1993 209 ' 0('174 . 0) 15 -3 r-i 00 (1930) 1 01 -16 90' -30 24 -3.'S6] . i993 0 .•is40" T Mehta1 .993' -13 40(1060) 70 (1360) .6 20 -9 70.. -IS 09 -4.31] 9 i 93 "0 :3 ' s9 4 • Subtotal 9 (5% CI): .7 21- -10 71 -is 70 -S.72]. • ' s s : , . Test for heterogenetiy:'«ChO.9Ei;f = 1 0 (P 3 . 3)=P = 0% Test for overal effect:42 .Z 0 »(P0.< 0001) Total 9 (5% CI) 719 1 , 00 00 -8 77 -101 . 1 -7.43] 85r Test for heterogenetiy: = Ch464 .2 df-13(P«0.0001),P 720 .% Test .for overal effect;12Z 8 . 2«(P «. 0.00001 ) 1 - 00. -50 ,0' 50 100 Favours Amo ldp in ie Favours Pa lcebo :  1  1  r  1  Figure 2. Example of a forest plot.  In a forest plot, the outcome of interest is listed at the top. The vertical line represents the line of no effect. Component studies are shown as squares (the size of which represents the weight) and the horizontal line running through each square shows the 95% confidence interval. The overall estimate is represented as a diamond; its centre represents the point estimate and the horizontal tips represent the confidence interval.  If the  confidence interval crosses the line o f no effect, there is no statistically significant difference between the intervention and the control. Data for changes i n blood pressure and heart rate were combined using a weighted mean difference method.  The withdrawals due to adverse events were analyzed using  relative risk, risk difference, and number needed to harm.  47  Tests for heterogeneity of treatment effect between the trials were performed using a standard chi-square statistic for heterogeneity. 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. Subgroup analyses were used to examine the results for specific categories of participants. The number of subgroup analyses performed should be kept to a minimum, since the greater number of hypotheses tested, the greater the number of differences one will find by chance alone (76). Possible subgroup analyses included: 1) Calcium channel blocker subclass: dihydropyridines, non-dihydropyridines (including benzothiazepines and phenylalkylamines) and others 2) Different formulations of the same active chemical entity 3) Age: children, adults, older people 4) Co-morbid conditions: ischemic heart disease, diabetes 5) Baseline severity of hypertension: mild, moderate, severe 6) Type of hypertension: elevated D B P , isolated systolic Sensitivity analyses are performed to assess how sensitive the results are to changes in the way the analyses were done (76). If these sensitivity analyses do not change the results substantially, more confidence can be placed in the results.  However, i f the results do  change, they must be interpreted with greater care. The robustness of the results of this systematic review were tested using the following sensitivity analyses, including: 1) Trials of high quality vs. poor quality 2) Trials that are industry-sponsored vs. non-industry sponsored 3) Trials with blood pressure data measured in the sitting position vs. other measurement positions  48  4)  Trials that used mercury sphygmomanometers vs. automatic devices  5) Trials with published standard deviations of blood pressure change vs. imputed standard deviations  2.2.10. Statistical Considerations 2.2.10.1.  Individual Study D a t a  In the RevMan program, continuous outcomes such as change in blood pressure and heart rate for each study must be inputted with group size, mean response, and standard deviation. For each study a weighted mean difference is calculated as W M D = m m . i r  Table 7: Data input for continuous data in RevMan Study i Group size Mean response Intervention m Control N , 2  2 i  Standard deviation sd,,sd,,  2;  For dichotomous (binary) outcomes, the number of subjects who experience an event in the intervention and control groups and the total number of subjects in each group are inputted in RevMan.  Table 8: Data input for dichotomous data in RevMan Study i Event N o event Intervention a, Control d, ;  Total n, N,.  c  2  Dichotomous data can be represented in four different ways in RevMan: Peto odds ratio, odds ratio, relative risk and risk difference (Table 9). Table 9: Formulae for individual study responses for dichotomous data in RevMan Estimate (for each study z) Formula OR, = exp{a,-E[a,]/v,}, where Peto odds ratio  Efa^n.^.+cVHOdds ratio Risk ratio or relative risk Risk difference  O R , = a,d,/b,c, RR,.= (a,./n„)/(c,/n ,) R D , = (a,/n„) - (c,/n ,) 2  2  49  2.2.10.2.  P o o l i n g trials  There are two main approaches for summarizing the results of studies in a metaanalysis: the "fixed effect" model and the "random effects" model.  The fixed effect  approach assumes that the group of studies being analyzed is a distinct population (a "fixed" set) and the aim is to estimate the mean effect size of these trials only. The fixed effect model tests the significance of the null hypothesis (i.e., no overall difference between treatment and control in the selected group of studies). Because the fixed effect approach ignores between-study variation, it assumes the studies have homogeneous effect sizes. The statistic j g is calculated to test for interstudy heterogeneity.  ff statistically significant  heterogeneity is present, the fixed-effect estimate of the overall effect size may not be valid (89).  The inverse variance method is used to combine weighted mean differences for  continuous data. The effect size for each study is weighted by the reciprocal of the square o f the standard error.  Thus, studies with large sample size are weighted more heavily than  smaller trials. Likewise, studies with smaller variances have greater influence on the overall effect size than studies with large variances.  For dichtomous data, the Mantel-Haenszel  method is used to produce odds ratio, relative risk, and risk difference estimates. The odds ratio and relative risk are relative measures and their values are similar i f the outcome is rare (76). However, the relative risk is recommended because the odds ratio is more difficult to interpret. Risk difference expresses results in absolute terms. The random effects model assumes that the set of studies being analyzed is a random sample from a larger population of studies. The DerSimonian and Laird method is used to combine continuous data or dichotomous data. The weight assigned to each study takes into account both intra- and inter-study variances. This approach accounts for interstudy  50  heterogeneity and produces wider confidence intervals than the fixed effect approach. A n advantage of the random effects approach is that it estimates the effect size for a hypothetical population of studies, which may include future studies or previous studies that may have been missed (89). Which of the two competing models is used depends on whether or not we wish to take into account between-study variation, f f there is no statistically significant heterogeneity between studies, the  fixed-effect  and random-effects  estimates will be similar.  A  disadvantage of the random effects model is that more weight is assigned to smaller studies compared to the fixed effect model. These small studies are more likely to be of poorer quality and are more subject to publication bias (76).  2.2.11 S t a r t i n g  doses  fn order to combine trial data in the analyses, calcium channel blocker drugs were analysed as increments of the recommended starting dose, as determined from reference pharmacopeias (44, 90). ff a range was given, the lower dose was taken as the starting dose. Table 10 lists starting doses of calcium channel blockers assessed in this review.  This  approach is based upon the assumption that starting doses recommended by companies must have evidence for blood pressure lowering efficacy and that the blood pressure lowering efficacy of each of the starting doses is approximately the same.  51  T a b l e 10: Starting doses of c a l c i u m channel blockers (44, 90) Drug B r a n d name (Company) Starting d o s e / d a y for hypertension Amlodipine Norvasc® (Pfizer) 5 mg Also available in combination with benazepril- Lotrel® (Novartis) Barnidipine Hypoca®, Vasexten® 10 mg (Yamanouchi) n/a Darodipine 100 mg* Diltiazem Felodipine Isradipine Lacidipine  Lercanidipine Lidoflazine Manidipine Nicardipine Nifedipine Nilvadipine Nisoldipine Nitrendipine Pranidipine Tiapamil  Verapamil  Cardizem®, Tiazac® (Biovail) Generic available Plendil® (AstraZeneca) Renedil® (Sanofi-Aventis) DynaCirc® (ReHant), Lomir® Lacipil® (GlaxoSiTUthKline), Caldine®, Lacimen®, Lacipil®, Midotens®, Motens® Zanidip® (Recordati) Clinium® (Janssen) Calslot® (Takeda) Cardene® (Roche) Adalat® (Bayer) Generic available Escor®, Nivadil® (Fujisawa) Sular® (Bayer) Bayotensin®, Baypress® (Bayer) n/a n/a  Isoptin® (Abbott) Generic available  Available i n Canada? Yes  No No  120-240 mg  Yes  5 mg  Yes  5mg 2-4mg  No No  10 mg Angina only 10 mg 60 mg 20-30 mg  No No No No Yes  8mg 10 mg 5-20 mg  No No No  2 mg* N o doses were statistically significantly different from placebo* 180-240 mg  No No  Yes  * F o r darodipine, pranidipine and tiapamil, no dosing information was available in the monographs; therefore dose Listed is the lowest effective dose as determined by the systematic review  2.2.12. Direct and indirect comparisons between doses When possible, direct and indirect comparisons of effect sizes between doses were performed  for each C C B drug.  In the direct method, only trials that randomized  participants to different doses were included in the analysis.  In the indirect method, an  52  "adjusted indirect comparison" and the associated standard error were calculated using the method described by Bucher et al (91) and Song et al (92). Briefly, C C B drug doses were compared from all the placebo-controlled trials, by adjusting for the common placebo intervention. The difference between the indirect and direct estimates was then assessed.  53  3.  RESULTS  3.1 Search findings T a b l e 11: Results of the search strategy N u m b e r of publications Publications identified by search strategy  4093  Trials excluded upon reading tides/abstracts  3076  Review articles identified by search strategy  563  Relevant review articles retrieved  138  Trials retrieved for detailed reading  454  Trials excluded upon detailed reading  217  Number of trials meeting inclusion criteria  237  O f the trials meeting the inclusion criteria: Data available  Included Studies  Data not available  Excluded Studies  106 131  The search strategy was highly sensitive but lacked specificity since 75% of hits (not including review articles) were excluded upon reading the tides or abstracts. There was an overlap of 983 studies between Medline and E M B A S E databases. One hundred thirty-eight relevant review articles identified by the search strategy were checked for references to trials. 102/106 (96%) of the included studies were found in the Cochrane Central Register of Controlled Trials ( C E N T R A L ) .  References that have not been indexed yet in C E N T R A L  have been sent to the Cochrane Hypertension Group for revision of the database. T w o of the missing studies (Circo 1997, Ninci 1997) were also not indexed in Medline or E M B A S E , but rather, were found from the reference list of a review article. Four of the 106 included studies were originally excluded due to lack of reporting of the number of patients in each arm; however these trials were included after obtaining the  54  missing values from Law et al's published meta-analysis of five classes of antihypertensives (93).  3.2 Characteristics ofincluded studies O f the 4093 citations identified by the search strategy, 106 trials (2.6%), published between 1976 and 2003, met the inclusion criteria and contained extractable data for 19 C C B drugs. O f these included trials, 99 (92%) of the studies were published in English, 3 (3%) in French, 3 (3%) in German, and 2 (2%) in Italian. One hundred four (97%) of the included studies had a parallel-group design, while only 3 (3%) were cross-over studies (only the precross-over data were used in this analysis).  Tables 12-30 summarize the characteristics of each included study. Each study was assigned a unique identifier consisting of either the trial acronym or the surname of the first author followed by the year of publication.  55  T a b l e 12: A m l o d i p i n e - Characteristics of included studies Study Chrysant 2003 (94)  Study Description Design: M C / R / D B / P C  trial.  Placebo run-in period: 4 weeks.  Treatment duration: 8 weeks. Country: U S A Quality: Cochrane method = B ; jadad score = 3 Participants: sitting D B P 100-115 m m H g and mean daytime D B P 90-119 mm H g via A B P M for inclusion into trial. N = 252: 161 m, 91 f. Mean age: 52 years (> 18 years for inclusion). Baseline B P : 155.1/103.8 m m Fig (Ami. group), and 154.2/103.3 mm H g (placebo group).  Baseline pulse pressure: 51.3 m m H g (Ami.  group), and 50.9 mm H g (placebo group). Interventions: Amlodipine 5 mg/day, olemesartan medoxomil 20 mg/day, or placebo. Primary and secondary outcomes: Change- from baseline in: mean 24-h D B P / S B P by A B P M (Instrument: not reported), sitting trough SBP, D B P by cuff (Instrument: not reported); response rates for: D B P < 90 mm H g and < 85 m m H g ; SBP < 140 mm H g and < 130 m m HgN o t e s : Change in sitting B P data, last observation carried forward, extracted from Table 3. SD of mean change in B P imputed from other trials. W D A E extracted from text, p. 427. F u n d i n g source: Sankyo Pharma Inc. Farsang 2001 (95)  Design: M C / R / D B / P C  trial.  Placebo-run-in period: 4 weeks.  Treatment duration: 8 weeks. Country: France, Flungary, Poland, South Africa, United Kingdom. Quality: Cochrane method =B; jadad score = 4 Participants: sitting D B P 95-114 mm Fig for inclusion into trial. N = 341: 231 m, 110 f. Mean age: 52.2 years (inclusion range: 20 - 80). Baseline BP: 164.8/102.5 m m H g (amlodipine group) and 161.5/102.1 mm  H g (placebo group).  Baseline pulse pressure: 62.3 m m H g  56  (amlodipine group) and 59.4 m m Fig (placebo group). Interventions: Amlodipine 5 mg/day, candesartan cilexetil 8 mg/day, cadesartan cilexetil 8 mg + amlodipine 5 mg/day combination, or placebo. Primary a n d secondary outcomes: Change from baseline in sitting and standing trough SBP, D B P and H R (Instrument: automatic device — Omron). N o t e s : Change in sitting B P and S D data extracted from Figure 1. W D A E reported. SDs of B P change were imputed from other trials because reported SDs were spuriously low. Author contacted — no response. F u n d i n g source: Astra Hassle A B , Sweden. Frick 1988 (96) [Duplicate publication: Frick 1989 (97)]  D e s i g n : M C / R / D B / P C dose-response study.  Washout period: 2  weeks. Placebo run-in period: 4 weeks. Treatment duration: 4 weeks, followed by dose titration in non-responders and an additional 4 weeks. Country: Finland Quality: Cochrane method =B; Jadad score — 4 Participants: supine and standing D B P 95-114 m m H g or 100-114 mm H g i f aged > 60 years for inclusion into trial. N = 205: 122 m, 83 f. Mean age: 50.2 years (range: 23-74). Baseline BP: 156.5/105.6 m m H g (Ami. 1.25 mg group), 154.8/106.3 m m H g (Ami. 2.5 mg group), 161.1/106.4 m m H g (Ami. 5 mg group), and 157.0/106.0 mm H g (placebo group).  Baseline pulse pressure: 50.9  m m H g (Ami. 1.25 mg group), 48.5 mm H g (Ami. 2.5 mg group), 54.7 m m H g (Ami. 5 mg group), and 51 mm H g (placebo group). Interventions: Amlodipine 1.25 mg/day, 2.5 mg/day, 5 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in supine and  standing  trough  SBP,  DBP  (Instrument:  standard  Hg  sphygmomanometer) and H R .  57  Notes: Change in standing B P data at week 4 extracted from Table 3 (Frick 1989). S D o f mean change in B P imputed from weighted mean SD o f other trials.  W D A E reported but cannot determine i f they  occurred before or after titration in non-responders. Pfizer (Kent, U K ) contacted — no data. Funding source: N o t reported. Frishman 1995 (98)  Design:  M C / R / D B / P C trial.  Washout period: 2 weeks.  Placebo  run-in period: 2-4 weeks. Treatment duration: 8 weeks.  Country: U S A Quality: Cochrane method = B; Jadad score = 4 Participants: Sitting D B P 100-115 m m Fig for inclusion into trial. N = 332: 209 m, 123 f. Mean age: 53.7 years (inclusion range: 21-80). Baseline BP: 157.8/103.8 m m H g (amlodipine group) and 158.0/103.9 mm H g (placebo group).  Baseline pulse pressure: 54.0 (amlodipine  group) and 54.1 (placebo group) Interventions:  Amlodipine  2.5 mg/day, benazepril 10 mg/day,  benazepril 10 mg + amlodipine 2.5 mg/day combination, or placebo. Primary and secondary outcomes: Change from baseline in sitting and  standing  trough  S B P , D B P (Instrument:  standard  Hg  sphygmomanometer), and H R . Notes: sitting B P data extracted from Figure 2. S D of mean change in B P imputed from SDs o f week-8 BPs. W D A E reported. Author contacted Funding source: Ciba-Geigy Corp. Kuschnir 1996 (99)  Design: M C / R / D B / P C trial. Washout period: 2 weeks. Placebo runin period: 2-4 weeks. Treatment duration: 8 weeks. Country: Argentina Quality: Cochrane method = B; Jadad score — 4 Participants: Sitting D B P 100-120 m m H g for inclusion into trial. N = 308: 138 m, 170 f. Mean age: 56.5 years (inclusion range 21-80). Baseline BP: 165.6/106.5 m m Fig (amlodipine group) and 166.4/106.9  58  mm  H g (placebo group).  Baseline pulse pressure: 59.1 m m H g  (amlodipine group) and 59.5 m m H g (placebo group). Interventions:  Amlodipine  5 mg/day,  benazepril  20  mg/day,  amlodipine 5 mg + benazepril 20 mg/day combination, or placebo. Primary and secondary outcomes: Change from baseline i n sitting trough SBP, D B P (Instrument: H g manometer), and H R . Notes: sitting B P data extracted from Table II. S D of mean change in SBPs imputed from S D of baseline BPs and S D of mean change in D B P imputed from other trials. W D A E reported. Attempted to contact author — no response. Funding source: Ciba Geigy Corp., U S A . Licata 1993 (100)  Design: R / D B / P C trial. Placebo run-in period: 2 weeks. Treatment duration: 4 weeks. Country: Italy Quality: Cochrane method — B; Jadad score = 3 Participants: Supine D B P >90 m m H g for inclusion into trial. N  = 30: 14 m, 16 f.  Mean age: 45.4 years (range: not reported).  Baseline B P : 165.9/100.0mm H g (treatment group) and 163.1/101.0 mm H g (placebo group).  Baseline pulse pressure: 65.9 (treatment  group) and 62.1 (placebo group). Interventions: Amlodipine 10 mg/day or placebo. Primary and secondary outcomes: Change from baseline in supine trough SBP, D B P , (Instrment: H g sphygmomanometer), and H R ; renal hemodynamic measurements. Notes: Supine B P data extracted from Table 1. S D of mean change in BPs imputed from SDs of week-4 BPs. W D A E not reported. Funding source: N o t reported. Mehta 1993 (101)  Design: M C / R / D B / P C dose-finding study. Placebo run-in period: 4 weeks. Treatment duration: 4 weeks.  Country: U S A Quality: Cochrane method = B; Jadad score = 4  59  Participants: Supine and standing D B P 95-114 m m H g for inclusion into trial. N = 210: 112 m, 98 f. Mean age: 52.8 years (inclusion range: 18-75). Baseline B P : 147.3/101.6 mm H g (Ami. 1.25 mg group), 150.7/102.2 mm H g (Ami. 2.5 mg group), 151.0/101/6 mm H g (Ami. 5 mg group), 152.8/101.9 (Ami. 10 mg group), and 153.4/102.0 m m H g (placebo group). Baseline pulse pressure: 45.7 m m H g (Ami. 1.25 mg group), 48.5 m m H g (Ami. 2.5 mg group), 49.4 (Ami. 5 mg group), 50.9 (Ami. 10 mg group), and 51.4 (placebo group). Interventions: Amlodipine 1.25 mg/day, 2.5 mg/day, 5 mg/day, 10 mg/day, or placebo. Primary a n d secondary outcomes: Change from baseline in supine and  standing  trough  sphygmomanometer),  SBP, D B P  (Instrument:  standard  Hg  and H R ; E C G ; body weight; biochemical,  hematologic, urinalysis lab tests. N o t e s : Change in standing B P and S E M extracted from Figure 1. W D A E reported. Author contacted — no data available. F u n d i n g source: Pfizer Research, U S A . Mroczek 1988 (102)  D e s i g n : R / D B / P C trial. Placebo run-in period: 4 weeks. Treatment  [Multiple publications: Mroczek 1991a (103), Mroczek 1991b (104), Burris 1994 (105)]  Country: U S A  duration: 4 weeks.  Quality: Cochrane method = B; Jadad score = 4 Participants: Standing and supine D B P 95-114 m m H g for inclusion into trial. N = 16: 11 m, 5 f.  Mean age: 56.9 years (range: 37-75). Baseline  standing BP: 154.9/102.2 m m H g (amlodipine group) and 146.9/105.6 m m Fig (placebo group).  Baseline pulse pressure: 52.7 (amlodipine  group) and 41.3 (placebo group). Interventions: Amlodipine 5 mg/day or placebo. Primary a n d secondary outcomes: Change from baseline in supine  60  and  standing  trough  SBP,  DBP  (Instrument:  standard  Hg  sphygmomanometer). N o t e s : Mean change in standing B P extracted from Table 2 (Mroczek 1991a) and S E M extracted from Figure 1 (Mroczek 1991b).  WDAE  reported. F u n d i n g source: Pfizer Inc., U S A . Pool 2001  D e s i g n : M C / R / D B / P C trial. Washout period: 2 weeks. Placebo run-  (106)  in period: 2-4 weeks. Treatment duration: 8 weeks. Country: U S A Quality: Cochrane method =A; Jadad score — 5 Participants: sitting D B P 100-115 m m H g for inclusion into trial. N = 454: 286 m, 168 f. Mean age: 53.8 years (inclusion range: 21-80). Baseline BP: 154.4/104.2 mm H g (amlodipine group) and 156.1/105.1 m m H g (placebo group).  Baseline pulse pressure: 50.2 (amlodipine  group) and 51 (placebo group). Interventions:  Amlodipine  5  mg/day,  benazepril  lOmg/day,  amlodipine 5 mg + benazepril 10 mg/day combination, or placebo. Primary and secondary outcomes: Change from baseline in sitting trough SBP, D B P (Instrument: not reported) and H R . N o t e s : Mean change in B P data extracted from Table 1. S D of mean change in BPs imputed from SDs from other trials. W D A E reported. Attempted to contact author — no response. F u n d i n g source: Novartis Pharmaceuticals, U S A .  61  Table 13: Barnidipine - Characteristics of included studies Study Hart 1997 (107)  Study Description Design: M C / R / D B / P C  trial.  Placebo run-in period: 4 weeks.  Treatment duration: 6 weeks, followed by dose titration in nonresponders and an additional 6 weeks. Country: Netherlands Quality: Cochrane method = B; Jadad score = 4 Participants: Sitting D B P 95-114 m m H g for inclusion into trial. N  = 190 (intention-to-treat): 124 m, 66 f.  Mean age: 55.8 years  (inclusion range: 18 -,75). Baseline B P (sitting): 161.0/102.0 mm H g (Bar. 10 mg/day group), 163.6/102.9 mm H g (Bar. 20 mg/day group), 166.2/102.5 m m H g (Bar. 30 mg/day group), and 165.2/102.5 m m H g (placebo group). Baseline pulse pressure: 59 mm H g (Bar. 10 mg/day group), 60.7 mm H g (Bar. 20 mg/day group), 63.7 mm H g (Bar. 30 mg/day group), and 62.7 m m H g (placebo group). Interventions:  Barnidipine 10 mg/day, 20 mg/day, 30 mg/day, or  placebo. Primary and secondary outcomes: Change from baseline in sitting and standing trough SBP, D B P (Instrument: sphygmomanometer) and HR. Notes: Change in sitting B P at week 6 extracted from Figure 1. SDs of changes in B P imputed from other trials. N o significant change in H R (no H R data provided at 6 weeks). W D A E not reported for each group separately. Attempted to contact author — no response. Funding source: Yamanouchi Europe B . V . , Netherlands.  62  Table 14: Darodipine - Characteristics of included studies Study Chrysant 1988 (108)  Study Description Design: R / D B / P C trial. Placebo run-in period: 3 weeks.  Treatment  duration: 4 weeks. Country: U S A Quality: Cochrane method = B ; Jadad score = 4 Participants: Supine D B P 100-115 m m H g for inclusion into trial. N = 43: 26 m, 17 f. Mean age: 51.7 years (range: not reported). Baseline B P (standing): 157/103 m m H g (Dar. 100 mg/day group), 160/104 m m H g (Dar. 200 mg/day group), 154/103 m m H g p a r . 300 mg/day group), and 157/102 m m H g (placebo group). Baseline pulse pressure: 54 m m H g (Dar. 100 mg/day group), 56 m m H g (Dar. 200 mg/day group), 51 m m H g (Dar. 300 mg/day), and 55 m m H g (placebo group). Interventions: Darodipine 100 mg/day, 200 mg/day, 300 mg/day, or placebo. Primary and secondary outcomes: Change from baseline i n supine and standing SBP, D B P (Instrument: not reported) and H R ; metabolic parameters ( B U N , glucose, electrolytes, Ever enzymes); E C G . Notes: standing B P and H R data and associated S E M s extracted from Figures 1-4. Changes in B P and H R calculated by subtracting baselines from weighted means of week 3 and 4 data. S D of change in B P and H R imputed from weighted mean S D of week 3 and 4 data. N o patients withdrew due to adverse events. Author contacted — data no longer available. Funding source: N o t reported.  63  Table 15: Diltiazem - Characteristics of included studies Study Burris 1990 (109)  Study Description Design: M C / R / D B / P C , 5 x 4 multifactorial-design trial.  Placebo  run-in: 4-6 weeks. Treatment duration: 6 weeks. Country: U S A Quality: Cochrane method = B; Jadad score — 3 Participants: Supine D B P 95-110 mm H g for inclusion into trial. N = 297: 184 m, 113 f. Mean age: 52.1 years (range: not reported). Baseline supine B P : 151.6/99.4 mm H g (all groups combined). Baseline pulse pressure: 52.2 mm H g (all groups combined). Interventions: Diltiazem SR 120 mg/day, 180 mg/day, 240 mg/day, 360 mg/day, hydrochlorothiazide 12.5 mg/day, 25 mg/day, 50 mg/day, all possible diltiazem SR-hydrochlorothiazide combinations thereof, or placebo. Primary and secondary outcomes: Change from baseline in supine and  standing  SBP,  DBP  (Instrument:  standard  Hg  sphygmomanometer) and H R ; % achieving goal B P (supine D B P < 90 mm H g at week 6); E C G ; lab tests. Notes: Change in D B P data extracted from Figure 1.  S D of B P  change imputed from other trials. Author contacted — no access to data; attempts to contact other authors/Biovail unsuccessful. Number of patients in each group was not explicidy reported; this data was obtained from a published metaanalysis (93). W D A E not reported separately for each group. Funding source: Marion Laboratories Inc. Chan 1997 (110)  Design: M C / R / D B / P C , 3 x 2 multifactorial-design trial.  Placebo  run-in: 4 weeks. Treatment duration: 12 weeks. Country: Taiwan and H o n g K o n g Quality: Cochrane method = B; Jadad score = 4 Participants: sitting D B P 95-114 mm H g for inclusion into trial.  \  N = 156: 134 m, 52 f. Mean age: 72 years (range: 65 - 88).  64  Baseline B P : 167.9/104.5 m m H g (Dilt. 120 mg group), 170.0/105.9 mm  H g (Dilt. 240 mg group) and 167.9/105.5 m m H g (placebo  group).  Baseline pulse pressure: 63.4 m m H g (Dilt. 120 mg group),  64.1 m m H g (Dilt. 240 mg group) and 62.4 m m H g (placebo group). Interventions: Diltiazem SR 120 mg/day, 240 mg/day, lisinopril 10 mg/day, diltiazem SR-lisinopril combinations 120/10 mg/day, 240/10 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in sitting SBP, D B P (Instrument: standard H g sphygmomanometer) and H R ; quality of life questionnaire. Notes: Change in B P data extracted from Table 2 and 3. Reported SDs of B P change were spuriously low (outside the 99% confidence interval of all trials reporting S D of B P change).  SDs were imputed  from endpoint BPs. W D A E reported. Author contacted: data no longer available and verification of data extraction not possible. Funding source: N o t reported. Cushman 1998 (111)  Design: M C / R / D B / P C trial. Washout period: 1 week. Placebo runin period: 4 weeks. Treatment duration: 12-week double-blind phase, followed by 36-week open label phase.  Country: U S A Quality: Cochrane method — B; Jadad score = 4 Participants: Sitting D B P 95-115 m m H g for inclusion into trial. N = 891: 609 m, 282 f. Mean age: 54.4 years (range: not reported). Baseline  B P : 152.6/101.7  mm  Hg  (Dilt.  120 mg/day  group),  155.4/102.2 mm H g (Dilt. 180 mg/day group), 155.4/101.7 m m H g (placebo group).  Baseline pulse pressure: 50.9 mm H g (Dilt. 120  mg/day group), 53.2 m m H g (Dilt. 180 mg/day group), 53.7 mm H g (placebo group). Interventions: Diltiazem E R 120 mg/day, 180 mg/day, enalapril 5 mg/day, enalapril + diltiazem E R combination 5/120 mg/day, 5/180  65  mg/day, or placebo. Primary and secondary outcomes: Change from baseline in sitting trough D B P , SBP (Instrument: standard H g sphygmomanometer) and H R ; trough-to-peak ratios. Notes: B P data at week 12 extracted from Table 2. S D of change in B P imputed from other trials. Author contacted: request for data was forwarded (no response). Funding source: Merck Research Laboratories, U S A . Djian 1990 (112)  Design: R / D B / P C dose-response trial. days.  Placebo run-in period: 15  Treatment duration: 4 weeks, followed by 4 weeks on forced  titration dose. Country: France Quality: Cochrane method = B; Jadad score = 3 Participants: Supine D B P 95-115 mm H g for inclusion into trial. N  = 95: 51 m, 44 f. Mean age: 58.4 years (range: not reported).  Baseline standing B P : 164.7/102.6 mm H g (Dil. 240 mg/day), 169.2/105.9 mm Fig (Dil. 300 mg/day), 167.9/103.8 m m H g (Dil. 360 mg/day), 168.9/107.1 mm H g (placebo group). Baseline pulse pressure: 62.1 mm H g (Dil. 240 mg/day group), 63.3 m m Fig (Dil. 300 mg/day group), 64.1 mm H g (Dil. 360 mg/day group), 61.8 m m H g (placebo group) Interventions: Diltiazem sustained release 240 mg/day, 300 mg/day, 360 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in SBP, D B P (Instrument: not reported) and H R , E C G (PR interval); lipid parameters. Notes: Reported variabilities assumed to be S E values.  Changes in  standing B P extracted from Table 2 and Figure 3. S D of change in standing D B P imputed from S E of change in supine D B P from text, p. S41. S D of change in SBP imputed from SDs of baseline SBP. H R data (from E C G s ) extracted from Table 3; S D of H R change imputed  66  from baseline SDs. Funding source: N o t reported. Felicetta 1992 (113)  Design: M C / R / D B / P C dose-response trial. Placebo run-in period: 46 weeks. Treatment duration: 4 weeks.  Country: U S A Quality: Cochrane method = B; Jadad score = 3 Participants: Supine D B P 95-110 m m H g for inclusion into trial. N = 229: 162 m, 67 f. Mean age: 48.8 years (inclusion range: 18-60). Baseline BP: 146.9/100.0 m m Fig (Dil. 90 mg/day group), 148.4/100.3 mm H g (Dil. 180 mg/day group), 150.9/99.9 m m H g (Dil. 360 mg/day group), 152.0/101.0 m m H g (Dil. 540 mg/day group), 153.3/100.4 mm H g (placebo group). Baseline pulse pressure: 46.9 m m H g (Dil. 90 mg/day group), 48.1 mm H g (Dil. 180 mg/day group), 51.0 m m H g (Dil. 360 mg/day group), 51 m m Fig (Dil. 540 mg/day group), 52.9 mm H g (placebo group). Interventions: Diltiazem C D 90 mg/day, 180 mg/day, 360 mg/day, 540 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in trough and peak supine SBP, D B P (Instrument: not reported), and H R ; % satisfactory B P response, defined as supine D B P < 90 m m H g or > 6 mm H g reduction; trough-peak ratios. Notes: Change in B P and associated S D extracted from Table II. W D A E reported. Funding source: N o t reported. Fiddes1994 (114)  Design: M C / R / D B / P C  trial.  Placebo run-in period: 4 weeks.  Treatment duration: 4 weeks, followed by titration in non-responders and an additional 8 weeks.  Country: U S A Quality: Cochrane method = B; Jadad score = 3 Participants: Supine D B P 95-114 m m H g for inclusion into trial. N = 350: 156 m, 194 f. Mean age: 64.7 years (> 55 for inclusion).  67  Baseline standing B P : 157.1/100 m m H g (diltiazem group) and 157.1/100 mm H g (placebo group). Baseline pulse pressure: 57.1 mm H g (diltiazem group) and 57.1 mm H g (placebo group). Interventions: Diltiazem (extended-release) 240 mg/day, or placebo. Primary a n d secondary outcomes: Change from baseline in trough supine/standing D B P , SBP, and H R ; % achieving B P < 90 m m H g ; lab tests. N o t e s : Change in supine D B P data at week 4 extracted from Figure 1. SD of D B P change imputed using pooled S D calculated from p value. W D A E reported but cannot determine i f they occurred before or after titration in non-responders. Attempted to contact author: no response. F u n d i n g source: N o t reported. Herpin 1990 (115)  D e s i g n : R / D B / P C trial. Placebo run-in period: 1 week.  Treatment  duration: 3 weeks. Country: France Quality: Cochrane method = B; Jadad score — 3 Participants: Supine D B P 95-115 m m Fig for inclusion into trial. N = 38 (efficacy analysis): 16 m, 22 f. Mean age: 51.4 years (inclusion range: 18 - 75).  Baseline B P : 164/101 mm H g (Dil. 240 mg/day  group), 163/102 mm Fig (Dil. 300 mg/day group), and 163/100 m m Fig (placebo group). Baseline pulse pressure: 63 m m H g (Dilt. 240 mg/day group), 61 m m H g (Dilt. 300 mg/day group), and 63 m m H g (placebo group). Interventions: Diltiazem SR 240 mg/day, 300 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in clinic supine D B P , SBP (Instrument:  H g manometer), and ambulatory  S B P / D B P (Instrument: Spacelab monitor). N o t e s : Clinic B P data extracted from Table II.  SDs of B P change  imputed from endpoint BPs. W D A E reported, but number of patients initially randomized to each group is not reported.  68  F u n d i n g source: N o t reported. Meeves 1994 (116)  D e s i g n : M C / R / D B / P C trial.  "Placebo run-in period: 4-6 weeks.  Treatment duration: 4 weeks (forced titration at week 1). Country: U S A Quality: Cochrane method = B; Jadad score = 3 Participants: Supine D B P 95-110 m m H g for inclusion into trial. N  = 111: 57 m, 54 f.  Mean age: 53.3 years (> 18 for inclusion).  Baseline supine B P : 149.5/99.7 m m H g (Diltiazem group), and 149.3/99.6 mm H g (placebo group). Baseline pulse pressure: 49.8 m m H g (Diltiazem group), and 49.7 mm H g (placebo group). Interventions: Diltiazem C D 180 mg/day (1 week)  300 mg/day (3  weeks), or placebo. Primary and secondary outcomes: Change from baseline in SBP, D B P (Instrument: standard H g sphygmomanometer), and H R ; troughpeak ratios; E C G . N o t e s : Change in B P , H R and associated S E data extracted from Table II. W D A E extracted from text, p. 234. F u n d i n g source: Marion Merrell D o w Inc., U S A . Pool 1993 (117)  D e s i g n : M C / R / D B / P C trial.  Placebo run-in period: 4-6 weeks.  Treatment duration: 6 weeks. Country: U S A Quality: Cochrane method = B; Jadad score = 4 Participants: Supine D B P 95-110 m m H g for inclusion into trial. N = 254: 168 m, 86 f. Mean age: 54.4 years (inclusion range: 18 -70). Baseline B P : 152.7/99.4 mm H g (diltiazem group) and 152.9/99.9 (placebo group). Baseline pulse pressure: 53.3 (diltiazem group) and 53.0 (placebo group). Interventions: Diltiazem SR 120 mg/day, hydrochlorthiazide 12.5 mg, combination diltiazem SR-hydrochlorthiazide 120/12.5 mg/day, or placebo.  69  Primary and secondary outcomes: Change from baseline in trough supine/standing SBP, D B P (Instrument: H g sphygmomanometer) and H R ; E C G ; fasting lab tests; 12-hour B P monitoring at baseline and week 6. Notes: Change in supine S B P / D B P and associated S E data extracted from Table 2. W D A E for each group not reported. Funding source: Marion Merrell D o w Inc., U S A . Prisant 2000 (119)  Design: M C / R / D B / P C trial with 3 x 4 factorial design. Placebo runin period: 4 weeks. Treatment duration: 6 weeks (forced titration at week 1.)  Country: U S A Quality: Cochrane method = B; Jadad score = 3 Participants: Supine D B P 95-114 m m H g for inclusion into trial. N = 329: 198 m, 131 f Mean age: 51.6 years (range: 19 - 86). Baseline  B P : 153.4/99.7  m m Fig (Dilt.  120 mg/day  group),  149.9/100.5 m m H g (Dilt. 180 mg/day group), 149.7/99.3 m m H g (Dilt. 240 mg/day group), 153.3/99.6 m m H g (Dilt. 360 mg/day group), and 150.7/99.8 m m H g (placebo group) Baseline pulse pressure: 53.7 m m H g (Dilt. 120 mg/day group), 49.4 m m H g (Dilt. 180 mg/day group), 50.4 m m H g (Dilt. 240 mg/day group), 53.7 m m Fig (Dilt. 360 mg/day group), and 50.9 m m H g (placebo group). Interventions: Diltiazem X R 120 mg/day, 180 mg/day, 240 mg/day, 360 mg/day, indapamide 1.25 mg/day, 2.5 mg/day, or diltiazem X R + indapamide combination 120/1.25, 180/1.25, 240/1.25, 120/2.5, 180/2.5, 240/2.5 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in supine clinic SBP, D B P (Instrument: not reported) and ambulatory B P (Instrument: SpaceLabs); hematologic, serum chemistry and urinalysis testing; E C G . Notes: Change in supine SBP and D B P extracted from Table 3. SDs  70  of B P change imputed from other trials. W D A E from each group not reported. F u n d i n g source: Rhone-Poulenc Rorer, Inc., U S A . Weir 1992 (119)  D e s i g n : M C / R / D B / P C trial.  Placebo run-in period: 4-6 weeks.  Treatment duration: 12 weeks (3 forced titration periods of 4 weeks each.) Country: U S A Quality: Cochrane method = B; Jadad score = 3 Participants: Supine D B P 95-110 m m H g for inclusion into trial. N - 298 (efficacy analysis N=274): 179 m, 95 f. Mean age: 53.5 years. Baseline BP: 150.1/99.8 m m H g (diltiazem group) and 152.7/99.5 m m H g (placebo group). Baseline pulse pressure: 50.3 (diltiazem group) and 53.2 (placebo group). Interventions: Diltiazem SR 60 mg/day (4 weeks)  90 mg/day (4  weeks) - ^ 1 2 0 mg/day (4 weeks); hydrochlorthiazide 6.25 (8 weeks) 12.5 mg/day (4 weeks); diltiazem SR/hydrochlorthiazide 60/6.25 mg/day (4 weeks) -> 90/6.25 mg/day (4 weeks) ~> 120/12.5 mg/day (4 weeks), or placebo. Primary  a n d secondary  outcomes:  Change  from  baseline in  supine/standing SBP, D B P (Instrument: H g sphygmomanometer) and H R ; % responders, defined as supine D B P < 90 m m H g and/or decrease > 10 m m H g . N o t e s : Change in supine D B P and SBP and associated S E data were extracted from Figure 3 and 4. F u n d i n g source: Marion Merrell D o w Inc., U S A . Whelton 1992 (120)  Design: M C / R / D B / P C  trial.  Placebo run-in period: 4 weeks.  Treatment duration: 4 weeks. Country: U S A Quality: Cochrane method = B; Jadad score = 3 Participants: Supine D B P 95-114 mm H g for inclusion into trial.  71  N = 275: 167 m, 108 f. Mean age: 51.9 years (inclusion range: 18 - 80). Baseline  B P : 152.2/100.4  m m H g (Dilt  120 mg/day  group),  150.7/100.8 m m H g (Dilt. 240 mg/day group), 148.9/99.2 m m H g (Dilt. 360 mg/day group), 152.0/99.8 m m H g (Dilt. 480 mg/day group), and 150.9/100.0 m m H g (placebo group). Baseline pulse pressure: 51.8 m m H g (Dilt. 120 mg/day group), 59.9 mm H g (Dilt. 240 mg/day group), 49.7 m m H g (Dilt. 360 mg/day group), 52.4 m m H g (Dilt. 480 mg/day group), and 50.9 m m H g (placebo group). Interventions: Diltiazem SR 120 mg/day, 240 mg/day, 360 mg/day, 480 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in supine trough SBP, D B P (Instrument: standard H g sphygmomanometer); 24hr A B P M (Instrument: Spacelabs 90202). Notes: Change in supine S B P / D B P data extracted from Table II. S D of B P change imputed from other trials. W D A E reported. Funding source: Rhone-Poulenc Rorer, Inc., U S A .  72  T a b l e 16: Felodipine - Characteristics of included studies Study  Study Description  Black 2001 (121)  Design: M C / R / D B / P C trial. Washout period: < 8 weeks. Placebo run-in period: 4 weeks.  Treatment duration: 8 weeks, followed by  titration in non-responders ^ total duration of follow-up: 52 weeks. -  Country: U S A  Quality: Cochrane method = B; Jadad score = 4 Participants: sitting SBP 140-159 m m H g and D B P < 90 mm Fig for inclusion into trial (isolated systolic hypertension). N = 171: 84 m, 87 f. Mean age: 66 years (> 55 for inclusion). Baseline B P : 149/83 m m H g (Fel. group), and 150/84 m m H g (placebo group). Baseline pulse pressure: 66 mm H g (Fel. group), and 66 m m H g (placebo group). Interventions: Felodipine E R 2.5 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in trough sitting SBP at week 52 (last observation carried forward), D B P (Instrument:  Hg  sphygmomanometer),  FIR,  quality  of  life  measurements, weight, E C G measurements. Notes: Change in SBP at week 8 data extracted from Figure 1. S D of SBP change imputed from other trials. W D A E reported but cannot determine i f they occurred before or after titration in non-responders. F u n d i n g source: AstraZeneca, L P , U S A .  Felodipine Co-op 1987 (122) [Duplication publication: Hamilton 1987a (123)]  Design: M C / R / D B / P C trial. Washout period: 4 weeks. Placebo runin period: 2 weeks. Treatment duration: 8 weeks. Country: U K  Quality: Cochrane method = B; Jadad score = 3 Participants: D B P 95-110 m m H g for inclusion into trial. N = 109: 59 m, 50 f. Mean age: 52 years (range: 20-70 for inclusion). Baseline B P : 166.8/104.4 mm H g (Fel. 5 m g / d group), 165.9/104.4 mm H g (Fel. 10 m g / d group), 171.2/105.4 m m H g (Fe. 20 m g / d  73  group), and 162.9/104.4 m m H g (placebo group).  Baseline pulse  pressure: 62.4 m m H g (Fel. 5 m g / d group), 61.5 m m H g (Fel. 10 m g / d group), 65.8 m m H g (Fel. 20 m g / d group), 58.5 m m H g (placebo group). Interventions:  Felodipine 5 mg/day,  10 mg/day,  20 mg/day  (administered in two divided doses/day), or placebo. Primary  and  supine/standing  secondary  outcomes:  B P , H R (Instrument:  Change  from  baseline in  Hawksley random  zero  sphygmomanometer), ankle circumference, bodyweight. Notes: standing B P / H R and S D data (12 hours postdose) from Week 8 extracted from Figure 4. W D A E extracted from Table V . Funding source: N o t reported. Fagan 1997 (124)  Design: M C / R / D B / P C study.  Placebo run-in period: 2-4 weeks.  Treatment duration: 9 weeks (dose titration at week 3 and 6 in nonresponders).  Country: U S A Quality: Cochrane method = B; Jadad score = 4 Participants: Sitting D B P 95-115 m m H g for inclusion into trial. N = 243: 163 m, 80 f. Mean age: 58.2 years (range: 26-83). Baseline B P : 159/102 m m H g (Fel. group) and 159/101 m m H g (placebo group). Baseline pulse pressure: 57 m m H g (Fel. group) and 58 m m Fig (placebo group). Interventions: Felodipine E R 2.5 mg/day or placebo. Primary and secondary outcomes: Change from baseline in sitting trough SBP, D B P (Instrument: standard H g sphygmomanometer), FIR, body weight. Notes: B P data at week 3 extracted from Table 2. SDs of B P change were imputed from p values.  Patients stratified by age: older (>65  years) vs. younger patients (< 60 years). W D A E reported but cannot determine i f they occurred before or after titration i n non-responders. Funding source: Merck, Sharp and Dohme, U S A .  74  Fetter 1994 (125)  Design: R / D B / P C trial. Placebo run-in period: 2 weeks. Treatment duration: 12 weeks. Country: Germany Quality: Cochrane method = B; Jadad "score = 3 Participants: WFIO stage I-II hypertension ( D B P > 115 m m H g for inclusion into trial. N = 71: m / f ratio not reported. Mean age: 47.5 years (range: 30 - 65). Baseline BP: 164/97 mm Fig (Fel. group) and 161/98 m m Fig (placebo group). Baseline pulse pressure: 67 m m H g (Fel. group) and 63 m m H g (placebo group). Interventions: Felodipine 10 mg/day or placebo. Primary and secondary outcomes: Blood flow velocity i n extra- and intracranial arteries, B P (Instrument: not reported). Notes: Published in German. BPs extracted from Table 1. SDs of B P change imputed from other trials. W D A E from each dosage group not reported. Funding source: Astra Chemicals GmbPI, Germany.  Gradman 1997 (126)  Design: M C / R / D B / P C factorial design study. Placebo run-in period: 4 weeks. Treatment duration: 8 weeks. Country: U S A Quality: Cochrane method = B; Jadad score = 3 Participants: Sitting D B P 95-115 m m H g for inclusion into trial. N  = 707: 457 m, 250 f.  Mean age: 53.5 years.  Baseline B P :  155.5/101.9 m m H g (all patients). Baseline pulse pressure: 53.6 m m HgInterventions: Felodipine E R 2.5 mg/day, 5 mg/day, 10 mg/day, enalapril 5 mg/day, 20 mg/day, felodipine E R + enalapril combination 2.5/5, 2.5/20, 5/5, 5/20, 10/5, 10/20 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in trough sitting D B P , SBP (Instrument: not reported - auscultatory method),  75  and H R . Notes: Change i n B P data extracted from Figure 1 and 2. SDs of change in S B P imputed from baseline S B P value; S D of change in D B P imputed from other trials. W D A E from each dosage group not reported. Attempted to contact author — no response. Funding source: Astra Merck, U S A . Kiesewetter 1994 (127)  Design: R / D B / P C trial. Placebo run-in period: 2 weeks. Treatment duration: 4 weeks. Country: Germany Quality: Cochrane method = B; Jadad score — 4 Participants: D B P 90-120 m m H g and SBP > 140 m m H g for inclusion into trial. N — 104: 74 m, 30 f. Mean age: 56.2 years (range: 45-65 for inclusion). Baseline B P : 154/94 m m H g (Fel. group) and 152/93 m m H g (placebo group). Baseline pulse pressure: 60 m m H g (Fel. group) and 59 m m Fig (placebo group). Interventions: Felodipine 10 mg/day or placebo. Primary  and  (Instrument:  secondary  outcomes:  sphygmomanometer)  and  Peak  supine  SBP,  DBP  H R ; haemorheological  parameters; urinalysis. Notes: BPs and H R s extracted from text, page 361. S D of B P change imputed from S D of week 4 BPs. Author contacted — dose and W D A E information obtained. Funding source: N o t reported. Liedholm 1989 (128)  Design: M C / R / D B / P C trial. Washout period: 4 weeks. Placebo runin period: 4 weeks. Treatment duration: 4 weeks. Country: Sweden Quality: Cochrane method — B; Jadad score = 4 Participants: Supine D B P 95-120 m m H g for inclusion into trial. N = 151: 147 m, 4 f. Median age: 51.8 years (range: 38-68). Baseline BP: 156.2/99.5 m m H g (Fel. 10 mg/day group), 157.6/101.4 m m H g  76  (Fel. 20 mg/day group), and 158.6/101.4 m m H g (placebo group). Baseline pulse pressure: 56.7 mm H g (Fel. 10 mg/day group), 56.2 m m H g (Fel. 20 mg/day group), and 57.2 m m H g (placebo group). Interventions: Felodipine E R 10 mg/day, 20 mg/day, or placebo. Primary a n d secondary outcomes: Change from baseline in peak and  trough  supine/standing  SBP,  DBP  (Instrument:  Hg  sphygmomanometer) and H R ; blood and urine lab tests. N o t e s : Change in trough supine B P , H R data and associated SDs extracted from Table 2. W D A E reported. F u n d i n g source: A B Hassle (subsidiary of A S T R A Pharmaceuticals), Sweden. Scholze 1999 (129)  D e s i g n : M C / R / D B / P C 3 x 4 factorial design, combination dosefinding study. Washout period: N o t reported. Treatment duration: 6 weeks. Country: Germany Quality: Cochrane method = B; Jadad score = 4 Participants: W H O class f-II (mild-to-moderate hypertension) for inclusion into trial. (range: 18-73).  N = 507: 327 m, 180 £  Mean age: 50.2 years  Baseline B P : 165.4/106.7 m m H g (Fel. 5 mg/day  group), 166.0/107.0 mm Fig (Fel. 10 mg/day group), and 166.6/107.3 m m Fig (placebo group). Baseline pulse pressure: 58.7 m m H g (Fel. 5 mg/day group), 59 mm Fig (Fel. 10 mg/day group), and 59.3 m m H g (placebo group). Interventions: Felodipine E R 5mg/day, 10 mg/day, ramipril 2.5 mg/day,  5  mg/day,  10  mg/day,  all  ramipril-felodipine  ER  combinations thereof, or placebo. Primary and secondary outcomes: Change from baseline in trough standing/supine SBP, D B P (Instrument: H g sphygmomanometer), and a composite supine-standing mean arterial pressure. N o t e s : Change in supine BPs and associated SEMs extracted from Table 1.  77  Author contacted - baseline BPs, number of patients in intention-totreat population, W D A E in each group obtained from Aventis Pharma. F u n d i n g source: Hoechst, Germany. van Ree 1996 (130)  Design: M C / R / D B / P C  study.  Placebo run-in period: 4 weeks.  Treatment duration: 6 weeks. Country: Netherlands Quality: Cochrane method = B; Jadad score = 4 Participants: Sitting D B P 100-115 m m H g and SBP 140-200 m m H g for inclusion into trial. N = 129: 38 m, 91 f. Mean age: 63 years (range: 50 — 80 for inclusion). Baseline B P : 175/104 mm H g (Fel. 2.5 mg/day group) m m H g , 173/105 mm H g (Fel. 5 mg/day group), 177/105 m m H g (placebo group). Baseline pulse pressure: 71 mm H g (Fel. 2.5 mg/day group), 68 mm H g (Fel. 5 mg/day group), 72 m m H g (placebo group). Interventions: Felodipine E R 2.5 mg/day, 5 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in trough sitting B P (Instrument: random zero sphygmomanometer — Hawksley), FIR; 24-hour A B P M . N o t e s : Change in B P and associated S D (at week 6) data extracted from Figure 1. W D A E extracted from text, p. 615. F u n d i n g source: Astra Pharmaceutica B V , Netherlands.  Weber 1994 (131)  Design: M C / R / D B / P C  study.  Placebo run-in period: 4 weeks.  Treatment duration: 8 weeks. Country: U S A Quality: Cochrane method = B; Jadad score = 4 Participants: Sitting D B P 95-115 m m H g for inclusion into trial. N = 286: 186 m, 100 f. Mean age: 54-55 years in all groups (range: 2475).  Baseline B P : 151.4/100.6 m m H g (Fel. 2.5 mg/day group),  147.9/100.6 mm H g (Fel. 5 mg/day group), 153.8/101.2 m m H g (Fel. 10 mg/day group), and 154.4/101.0 m m H g (placebo group).  78  Baseline pulse pressure: 50.8 mm H g (Fel. 2.5 mg/day group), 47.3 m m H g (Fel. 5 mg/day group), 52.6 mm H g (Fel. 10 mg/day), and 53.4 m m H g (placebo group). Interventions: Felodipine E R 2.5 mg/day, 5 mg/day, 10 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in peak and  trough  sitting  DBP,  SBP,  (Instrument:  standard  Fig  sphygmomanometer), and H R . N o t e s : Change in trough sitting B P extracted from Table 1. S D of B P change calculated as weighted mean of SD of B P change at week 4 and 8 (extracted from Figure 1 and 2). W D A E extracted fom text, p.348. F u n d i n g source: Merck Research Laboratories, U S A . Wester 1991 (132) [Duplicate publication: Wester 1 9 9 0 abstract only (133)]  Design: M C / R / D B / P C  study.  Placebo run-in period: 4 weeks.  Treatment duration: 4 weeks. Country: U K and Netherlands Quality: Cochrane method = B; Jadad score = 4 Participants: Supine D B P 95-120 m m H g for inclusion into trial. N = 183: 88 m, 95 f.  Mean age: 53 years (range: 28-65). Baseline  standing BP: 161/105 m m H g (Fel. 5 mg/day group), 158/103 m m H g (Fel. 10 mg/day group), 164/106 mm H g (Fel. 20 mg/day group), and 165/103 mm Fig (placebo group). Baseline pulse pressure: 56 mm H g (Fel. 5 mg/day group), 55 m m H g (Fel. 10 mg/day group), 58 mm H g (Fel. 20 mg/day group), and 62 m m H g (placebo group). Interventions: Felodipine E R 5 mg/day, 10 mg/day, 20 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in trough supine  and  standing  SBP, D B P  (Instrument:  random  zero  sphygmomanometer - Hawksley), and H R . N o t e s : Standing BPs extracted from Table 1. S D of change in SBP imputed from week 4 SBP values.  S D of change in standing D B P  79  imputed from S D of change in supine D B P given in Figure 1. W D A E extracted from text, p. 279.  Funding source: Not reported.  80  T a b l e 17: Isradipine - Characteristics of i n c l u d e d studies Arosio 1993 (134)  Design:  R / D B / P C crossover study.  Washout period: at least 3  weeks. Treatment duration: 2 months (each treatment period was 1 month). Country: Italy Quality: Cochrane method = B; Jadad score = 2 Participants: "mild to moderate essential hypertension" for inclusion into trial.  N =16: 12 m, 4f.  Mean age: 41 years (range: 30-45).  Baseline B P : 157/102 m m H g (Israd. group) and 156/101 mm H g (placebo group).  Baseline pulse pressure: 55 (Israd. group) and 55  (placebo group). Interventions: Isradipine SR 5 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in supine SBP, D B P , H R (Instrument: automatic device - Dinamap); brachial artery hemodynamics. N o t e s : Pre-crossover data B P data (at week 4) extracted from Table 1. SDs of change in B P imputed from week 4 SDs of B P values. H R data and W D A E not reported. F u n d i n g source: N o t reported. Burger 1993 (135)  D e s i g n : M C / R / D B / P C trial.  Placebo run-in: 2 weeks.  Treatment  duration: 3 weeks. Country: Germany Quality: Cochrane method — B ; Jadad score = 3 Participants: D B P 100-115 mm H g for inclusion into trial. N = 37: 11 m, 18 f. Mean age: 56 years (range: 42 - 75). Baseline B P : 171/110 mm H g (Israd. group), 163/106 mm H g (placebo group). Baseline pulse pressure: 61 m m H g (Israd. group), and 57 mm H g (placebo group). Interventions: Isradipine S R O 10 mg/day, or placebo. Primary a n d secondary outcomes: Change from baseline in SBP, DBP  (Instrument:  sphygmomanometer), H R ; A B P M  (Instrument:  81  SpaceLabs). N o t e s : Published in German.  Office BPs and H R s extracted from  Table 2. SDs of B P / H R change imputed from other trials. F u n d i n g source: N o t reported. Chyrsant 1995a (136) [Duplicate publication: Chrysant 1995b (137)]  Design: M C / R / D B / P C Treatment  trial.  Placebo run-in period: 3 weeks.  duration: 6 weeks (weekly forced titration in 5 mg  increments to target dose in active treatment groups). Country: U S A Quality: Cochrane method = B; Jadad score = 3 Participants: Sitting D B P 100-114 m m H g for inclusion into trial. N = 402 (384 Intent-to-treat): approx. 75% "white males". Mean age: 56 (range: not reported). Baseline BP: 161/104 mm Fig (5 mg group), 158/104 mm H g (10 mg group), 156/104 (15 mg group), 156/104 (20 mg group), 158/104 m m H g (placebo group). Baseline pulse pressure: 57 (5 mg group), 54 (10 mg group), 52 (15 mg group), 52 (20 mg group), and 54 (placebo group). Interventions: Isradipine C R 5 mg/day (6 weeks), 10 mg/day (5 weeks), 15 mg/day (4 weeks), 20 mg/day (3 weeks), or placebo. Primary  a n d secondary  outcomes:  Change  sitting/supine/standing SBP, D B P (Instrument:  from  baseline in  N o t reported  -  auscultatory method), and FIR. N o t e s : Change in B P data extracted from Figure 1. S D of changes in S B P / D B P not available and were imputed using data from other trials. W D A E in each group not reported. Author contacted: baseline BPs obtained. F u n d i n g source: Sandoz Research Institute, U S A . Holmes 1993 (138)  Design: M C / R / D B / P C  trial.  Placebo run-in period: 3 weeks.  Treatment duration: 4 weeks. Country: N o t reported. Quality: Cochrane method = B; Jadad score = 3  82  Participants: Supine D B P 100-120 m m H g for inclusion into trial. N = 190: 84 m, 106 f. Mean age: 57 years (range: 21 - 89). Baseline B P : 176/108 m m H g (Israd. 2.5 mg/day group), 171/107 m m H g (Israd. 5 mg/day group), 172/107 mm H g (placebo group). Baseline pulse pressure: 68 m m H g (Israd. 2.5 mg/day group), 64 m m H g (Israd. 5 mg/day group) and 65 m m H g (placebo group). Interventions: Isradipine S R O 2.5 mg/day, 5 mg/day, or placebo. Primary and secondary outcomes: Change from, baseline in sitting SBP, D B P (Instrument: H g sphymomanometer), and H R — trough. N o t e s : unpublished B P / H R / W D A E data extracted from Sandoz archive files. Author contacted - B P / H R / W D A E data received. F u n d i n g source: Sandoz Pharma, Switzerland. Italian-Belgian Group 1989 (139)  D e s i g n : M C / R / D B / P C trial.  Placebo run-in: 3 weeks.  Treatment  duration: 5 weeks. Country: Italy and Belgium Quality: Cochrane method = B; Jadad score = 3 Participants: supine D B P 100-120 m m H g for inclusion into trial. N = 178: 85 m, 93 f. Mean age: 60 years (inclusion range: 30 - 85). Baseline B P : 171/104 mm H g (Israd. 1 mg/day group), 175/105 m m Tig (Israd. 2.5 mg/day group), 175/105 mm H g (Israd. 5 mg/day group), 176/104 m m H g (placebo group). Baseline pulse pressure: 67 mm H g (Israd. 1 mg/day group), 70 m m H g (Israd 2.5 mg/day group), 70 mm H g (Israd. 5 mg/day group), and 72 mm H g (placebo group). Interventions: Isradipine 1 mg/day, 2.5 mg/day, 5 mg/day, or placebo. Primary a n d secondary outcomes: Change from baseline in supine and standing SBP, D B P (Instrument: Tig sphygmomanometer), H R , N o t e s : Trough B P s / H R s and associated SDs extracted from Table II. SDs of SBP change and FIR change imputed from baseline SDs. S D of D B P change imputed from other trials. W D A E extracted from text,  83  p. 97. F u n d i n g source: N o t reported. Kirch 1990 (140)  D e s i g n : M C / R / D B / P C trial.  Placebo run-in: 2 weeks.  Treatment  duration: 4 weeks. Country: Germany Quality: Cochrane method = B; Jadad score — 4 Participants: Sitting D B P > 105 m m H g for inclusion into trial. N = 86: 46 m, 37 f (3 patients unaccounted for in demographics table). Mean age: 58.2 years (> 18 for inclusion). Baseline BP: 174/109 mm H g (Israd. 2.5 mg/day group), 175/108 mm H g (Israd. 5 mg/day group), 175/108 (Israd. 10 mg/day group), 176/108 m m H g (placebo group). Baseline pulse pressure: 65 mm H g (Israd. 2.5 mg/day group), 67 mm H g (Israd. 5 mg/day group), 67 m m H g (Israd. 10 mg/day group) and 68 m m H g (placebo group). Interventions: Isradipine 2.5 mg/day, 5 mg/day, 10 mg/day, or placebo. Primary a n d secondary outcomes: Change from baseline in sitting and standing SBP, D B P (Instrument: sphygmomanometer), and H R (2-3 hours post-dose). N o t e s : sitting BPs extracted from Table 2. S D of B P change imputed from weighted mean S D of week 3 and 4 BPs. W D A E extracted from text, p. S56. F u n d i n g source: N o t reported.  Man in't Veld 1991 (141)  D e s i g n : M C / R / D B / P C trial. Placebo run-in: 3-5 weeks. Treatment duration: 6 weeks. Country: N o t reported. Author from Netherlands Quality: Cochrane method = B ; Jadad score = 3 Participants: sitting D B P 95-1 f 5 mm Fig for inclusion into trial. N = 187: 99 m, 88 f. 110 patients received isradipine in the 4-week selection phase; the remaining 77 patients had received isradipine for up to 127 weeks before entering study. Mean age: 49 years (range: 23 -  84  65). Baseline B P : 158/102 m m H g (Israd. group), 160/102 m m H g (placebo group). Baseline pulse pressure: 56 m m H g (Israd. group), and 58 m m H g (placebo group). Interventions: Isradipine 5 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in sitting SBP, D B P (Instrument:  H g sphygmomanometer), responder rate  (DBP < 90 m m H g or fall of > 10 m m Hg). Notes:  A l l patients had demonstrated prior reponse to isradipine  treatment prior to trial entry. BPs extracted from Table 1. SDs of B P change imputed from weighted mean of SDs of week 4 and 6 BPs. W D A E extracted from text, p. 133S. F u n d i n g source: N o t reported. O'Grady 1997 (142)  Design: R / D B / P C  trial.  Placebo run-in: 2 weeks.  Treatment  duration: 4 weeks. Country: Austria Quality: Cochrane method = B ; Jadad score = 3 Participants: "mild to moderate hypertension", scintigraphalfy-visible atherosclerotic lesions of carotid vessels and platelet uptake ratio >1.15 for inclusion into trial. N = 40: 25 m, 15 f. Mean age: 45 years (range: 28 - 68). Baseline B P : 134/100 mm H g (group), 136/100 mm H g (placebo group). Baseline pulse pressure: 34 m m H g (Israd. group), and 36 mm Fig (placebo group). Interventions: Isradipine S R O 5 mg/day, or placebo. Primary a n d secondary outcomes: Platelet parameters (platelet uptake ratio, platelet survival), change from baseline in SBP, D B P (Instrument: not reported), and H R . N o t e s : BPs extracted from Fig. 1. SDs of B P change imputed from SDs of week 6 BPs. W D A E extracted from text, p. 367. F u n d i n g source: N o t reported.  "85  Pittrow 1997 (143)  D e s i g n : M C / R / D B / P C trial. Washout period: 2 weeks. Placebo runin: 2 weeks. Treatment duration: 6 weeks, followed by dose titration in non-responders and an additional 6 weeks. Country: Germany Quality: Cochrane method = B; Jadad score = 4 Participants: sitting D B P 100-114 mm H g for inclusion into trial. N = 405: 267 m, 138 f. Mean age: 55 years (range: 26 - 86). Baseline B P : 157.4/104.2  mm H g (Israd. 2.5  mg/day  group),  159.2/106.0 m m H g (Israd. 5 mg/day group), 150.4/104.1 mm H g (placebo group).  Baseline pulse pressure: 53.2 m m H g (Israd. 2.5  mg/day group), 53.2 m m Pig (Israd. 5 mg/day group), and 46.3 mm Pig (placebo group). Interventions: Isradipine S R O 2.5 mg/day, 5 mg/day; sprirapril 3 mg/day, 6 mg/day; isradipine S R O 2.5 mg + spirapril 3 mg/day combination, or placebo. Primary a n d secondary outcomes: Change from baseline in sitting SBP, D B P (Instrument: sphygmomanometer), FIR — trough and peak; standard lab tests; E C G . N o t e s : Change in trough BPs and associated SDs at week 6 extracted from Table 2A. n values extracted from Table 3. W D A E extracted from Table 1. F u n d i n g source: Sandoz, Germany. Prisant 1991 (144)  D e s i g n : M C / R / D B / P C trial.  Placebo run-in: 3 weeks.  Treatment  duration: 5 weeks. Country: U S A Quality: Cochrane method = B; Jadad score = 2 Participants: supine D B P 100-119 mm H g for inclusion into trial. N = 203 (170 in efficacy analysis): 117 m, 53 f. Mean age: 52.1 years (range: 22 - 77). Baseline B P : 156/104 trim Fig (Israd. 5 mg/day group), 160/104 m m Fig (Israd. 10 mg/day group), 157/103 m m H g (Israd. 15 mg/day  86  group, 152/104 m m H g (Israd. 20 mg/day group), 150/104 m m H g (placebo group). Baseline pulse pressure: 52 m m H g (Israd 5 mg/day group), 56 m m H g (Israd. 10 mg/day group), 54 m m H g (Israd. 15 mg/day group), 49 m m H g (Israd. 20 mg/day group), and 47 m m H g (placebo group). Interventions: Isradipine 5 mg/day (5 weeks), 10 mg/day (1 week at 5 mg/day  4 weeks at 10 mg/day), 15 mg/day (1 week at 5mg/day  1 week at 10 mg/day 10 mg/day  3 weeks at 15 mg/day), 20 mg/day (1 week at  1 week at 15 mg/day  3 weeks at 20 mg/day), or  placebo. Primary and secondary outcomes: Change from baseline in SBP, D B P (Instrument: not reported); E C G markers of ischemia. Notes: Peak BPs, H R s , and associated SDs extracted from Table II. SDs o f B P change and H R change imputed from, endpoint B P s / H R s . Funding source: N o t reported. Youssef 1992 (145)  Design: R / D B / P C trial. Treatment duration: 8 weeks.  [Duplicate publication: Youssef 1993 (146)]  Quality: Cochrane method = B ; Jadad score = 2  Country: Egypt  Participants: supine D B P 95-114 m m H g for inclusion into trial. N = 90: 57 m, 33 f. Mean age: not reported (range: 44 - 68). Baseline B P : 173/104 m m H g (Israd. group), 178/104 m m H g (placebo group). Baseline pulse pressure: 69 m m H g (Israd. group), and 74 m m H g (placebo group). Interventions: Isradipine 5 mg/day, enalapril 20 mg/day, benazepril 10 mg/day, xipamide 10 mg + triamterine 30 mg/day combination, or placebo. Primary and secondary outcomes: Change from baseline in supine SBP, D B P (Instrument: not reported); serum lipid profile. Notes: BPs extracted from Table 1 o f Youssef 1993 paper. SDs o f B P change imputed from other trials. Reported SDs o f B P at baseline and endpoint were spuriously low.  87  Attempts to contact author unsuccessful. F u n d i n g source: N o t reported.  T a b l e 18: L a c i d i p i n e - Characteristics of i n c l u d e d studies Study Rizzini 1991 (147)  Study Characteristics Design: M C / R / D B / P C  trial.  Placebo run-in period: 4 weeks.  Treatment duration: 4 weeks, followed by dose titration in nonresponders for 4 weeks, then 11 months open-label. Country: Italy Quality: Cochrane method = B; Jadad score = 3 Participants: Sitting D B P 95-115 mm H g and SBP < 200 m m H g for inclusion into trial. N = 131: 48 m, 83 f. Mean age: 70.3 years (range: 65 - 86). Baseline sitting B P : 175.5/101.2 mm H g (Lac. 2 mg/day group), 179.2/102.1 m m H g (Lac. 4 mg/day group), and 177.9/102.3 m m H g (placebo group). Baseline pulse pressure: 74.3 m m H g (Lac. 2 mg/day group), 77.1 mm H g (Lac. 4 mg/day group), and 75.6 mm H g (placebo group). Interventions: Lacidipine 2 mg/day, 4 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in sitting trough SBP, D B P (Instrument: H g sphygmomanometer), and H R . N o t e s : Mean change in sitting D B P at week 4 and associated confidence intervals extracted from Table 4. Mean change in sitting SBP at week 4 extracted from text on page S4f. S D of SBP change imputed from SDs of week 4 BP's. H R data and W D A E at week 4 not reported. F u n d i n g source: N o t reported.  89  T a b l e 19: L e r c a n i d i p i n e - Chatacteristics of included studies Study  Study Characteristics  Barbagallo 2000 (148)  D e s i g n : M C / R / D B / P C trial. Washout period: 1 week. Placebo runin period: 3 weeks. Treatment duration: 4 weeks, followed by .4 weeks of doubled dose in non-responders. Country: Italy Quality: Cochrane method = B ; Jadad score = 3 Participants: Elderly I S H patients.  SBP 160-220 m m H g and D B P  <95 m m H g for inclusion into trial. N = 83: 38 m, 45 f. Mean age: 66.7 years (> 60 for inclusion). Baseline B P : 172.6/87.1 m m H g (Lercan. group), 172.4/87.1 mm H g (placebo group). Baseline pulse pressure: 85.5 m m H g (Lercan. group), and 85.3 mm H g (placebo group). Interventions: Lercanidipine 10 mg/day, or placebo. Primary a n d secondary outcomes: Change from baseline in SBP, D B P (Instrument: not reported), and H R ; % responders, defined as decrease of SBP > 20 m m H g ; % patients with normalized SBP (SBP < 140 m m Hg); E C G ; lab tests. N o t e s : supine BPs and H R s in per protocol patients (n=70) extracted from Table 2. SDs of B P / H R change imputed from week 4 B P s / H R s . W D A E extracted from text, p. 378. Author contacted: responded with data clarification. F u n d i n g source: N o t reported. G r c o 1997 (149)  D e s i g n : M C / R / D B / P C trial. Washout period: 2 weeks. Placebo runin period: 3 weeks. Treatment duration: 4 weeks, followed by 12 weeks of dose titration in non-responders. Country: Italy Quality: Cochrane method = B; Jadad score — 2 Participants: D B P 95-115 m m H g for inclusion into trial. N = 132: 58 m, 74 f. Mean age: 54.7 years (range: 18-70). Baseline BP/pulse pressure: N o t reported.  90  Interventions: Lercanidipine 10 mg/day, 20 mg/day, or placebo. Primary a n d secondary outcomes: Change from baseline in sitting SBP, D B P (Instrument: not reported) and H R ; % responders (decrease in D B P > 10 m m Hg); % with normalized D B P ( D B P < 90 m m Hg); trough-to-peak ratios. N o t e s : N o baseline BPs provided and thus, no B P data was extracted. Trough H R data extracted from Table 3. SDs of H R change imputed from week 4 HRs. W D A E for each group not reported. Article not indexed in electronic databases — retrieved from reference list of review article. F u n d i n g source: N o t reported. N i n c i 1997 (150)  D e s i g n : M C / R / D B / P C trial. Washout period: 2 weeks. Placebo runin period: 3 weeks. Treatment duration: 4 weeks, followed by 12 weeks of dose titration in non-responders. Country: ftaly Quality: Cochrane method = B ; Jadad score = 3 Participants: Supine D B P 95-115 mm H g for inclusion into trial. N = 144: 77 m, 67 f. Mean age: 68.4 years (range: 60 - 85). Baseline B P : 171.1/101.6 mm H g g_ercan.group), 168.3/101.6 m m H g (placebo group). Baseline pulse pressure: 69.5 m m H g (Lercan. group), and 66.7 m m H g (placebo group). Interventions: Lercanidipine tO mg/day, or placebo. Primary and secondary outcomes: Change from baseline in sitting SBP, D B P (Instrument: not reported), and H R ; % responders ( D B P < 90 mm Hg); trough-peak ratios; E C G ; lab tests. N o t e s : Change in B P / H R extracted from Fig. f and 2, and text, p. S41.  SDs of B P / P f R change imputed from other trials.  WDAE  reported, but did not specify i f these occurred during the 1" 4 weeks. Article not indexed in electronic databases - retrieved from reference list of review article. F u n d i n g source: N o t reported.  91  Omboni 1998 (151)  D e s i g n : M C / R / D B / P C trial. Washout period: 2 weeks. Placebo runin period: 3 weeks. Treatment duration: 4 weeks. Country: Italy Quality: Cochrane method = B; Jadad score = 3 Participants: Supine D B P 90-109 m m H g for inclusion into trial. N  = 243: 156 m, 87 f.  Mean age: 51 years (range: not reported).  Baseline supine B P : 154/98 m m H g (Lercan. 2.5 mg/day group), 155/99 mm H g (Lercan. 5 mg/day group), 156/99 m m H g (Lercan. 10 mg/day group), 155/99 mm H g (placebo group).  Baseline pulse  pressure: 56 mm H g (Lercan. 2.5 mg/day group), 56 mm H g (Lercan. 5 mg/day group), 57 mm H g (Lercan. 10 mg/day group), and 56 m m H g (placebo group). Interventions: Lercanidipine 2.5 mg/day, 5 mg/day, 10 mg/day, or placebo. Primary a n d secondary outcomes: Change from baseline in supine SBP, D B P (Instrument: standard H g sphygmomanometer) and H R ; % responders, defined as D B P < 85 m m H g or D B P reduction > 10 m m Hg; 24-hr A B P M (Instrument: Spacelabs 90207). N o t e s : Change in B P and associated SDs extracted from Fig. 1. W D A E extracted from text, p. 1836. F u n d i n g source: Recordati S.p.A., Italy. Rimoldi 1993 (152)  D e s i g n : M C / R / D B / P C trial. Washout period: 1 week. Placebo runin period: 3 weeks. Treatment duration: 4 weeks. Country: Italy Quality: Cochrane method = B ; Jadad score = 1 Participants: Supine D B P 95-114 m m H g for inclusion into trial. N = 152: 67 m, 85 f. Mean age: 56.3 years (range: 21 - 70). Baseline D B P : 101.3 mm H g (Lercan. tab. group), 101.5 (Lercan. capsule group), 101.3 m m H g (placebo group).  Baseline SBP/pulse  pressure: not reported/cannot be calculated. Interventions:  Lercanidipine tablets 10 mg/day (1 week)  20  92  mg/day (3 weeks), lercanidipine capsules (same doses as tablets), or placebo. Primary a n d secondary outcomes: Change from baseline in supine and standing SBP, D B P (Instrument: not reported — by auscultation), H R — trough; E C G ; laboratory parameters. N o t e s : Tablet and capsule groups combined for RevMan entry. D B P s and H R s extracted from Table III. S D of D B P / H R change imputed from endpoint values. SBPs and W D A E not reported. F u n d i n g source: N o t reported. Rimoldi 1994 (153)  D e s i g n : R / D B / P C trial. Washout period: 2 weeks. Placebo run-in period: 3 weeks. Treatment duradon: 4 weeks. Country: Italy Quality: Cochrane method = B; Jadad score = 3 Participants: Supine D B P 95-114 mm H g for inclusion into trial. N = 30: 18 m, 12 f. Mean age: 50.8 years (range: not reported). Baseline B P : 148.9/103.3 mm H g (Lercan. fO mg/day group), 156.5/103.1 m m H g (Lercan. 20 mg/day), and 156.5/101.9 m m H g (placebo group).  Baseline pulse pressure: 45.5 m m H g (Lercan. 10  mg/day group), 53.5 m m H g (Lercan. 20 mg/day group), and 54.6 mm H g (placebo group). Interventions: Lercanidipine 10 mg/day, 20 mg/day, or placebo. Primary a n d secondary outcomes: Change from baseline in trough (mean  of 2 supine  and  f  standing measurement)  SBP, D B P  (fnstrument: not reported — by auscultation) and FfR; 24-hr A B P M (fnstrument: Spacelab Kontron). N o t e s : B P and H R data extracted from Table II. change imputed from SDs of week 4 B P s / H R s .  S D of B P / H R W D A E extracted  from text, p. 26. F u n d i n g source: N o t reported.  93  T a b l e 20: L i d o f l a z i n e - Characteristics of included studies Study MeilinkHoedemaker 1976 (154)  Study Characteristics D e s i g n : M C / D B / P C / c r o s s - o v e r trial. Treatment duration: 12 weeks per treatment sequence Country: Italy Quality: Cochrane method = A ; Jadad score = 3 Participants: Post-infarction patients. N = 10: 10 m, 0 f. Mean age: 43.8 years (range: 24 -58). Baseline B P : 139/96 m m H g (lidoflazine group), 137/94 m m H g (placebo group).  Baseline pulse pressure: 43 m m H g (lidoflazine  group), and 43 mm H g (placebo group). Interventions: Lidoflazine 180 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in sitting SBP, D B P (Instrument: H g sphygmomanometer), and H R (by E C G ) . N o t e s : Individual patient data reported.  B P data for patients with  baseline D B P > 90 m m H g and/or SBP > 140 m m H g for the first 12 weeks (before cross-over) extracted from Table II. D i d not explicitiy state trial was randomized. F u n d i n g source: N o t reported.  94  T a b l e 21: M a n i d i p i n e - Characteristics of i n c l u d e d studies Study Fogari 1996 (155)  Study Description D e s i g n : R / D B / P C trial. Placebo run-in period: 2 weeks. Treatment duration: 4 weeks. Country: Italy . Quality: Cochrane method = B; Jadad score = 3 Participants: Supine D B P 95-115 m m H g for inclusion into trial. N = 52: 32 m, 20 f. Mean age: 53.4 years (range: 40 - 63). Baseline B P : 163.8/104.7 m m H g (Man. 10 mg/day group), 166.0/104.7 m m H g (Man. 20 mg/day group), 164.9/105.3 mm H g (Man. 40 mg/day group) and 162/100 m m Fig (placebo group). Baseline pulse pressure: 59.1 m m H g (Man. 10 mg/day group), 61.3 mm H g (Man. 20 mg/day group), 59.6 m m H g , (Man. 40 mg/day group), and 62 m m H g (placebo group). Interventions: Manidipine 10 mg/day, 20 mg/day, 40 mg/day, or placebo. Primary  and secondary  outcomes:  Change  from  baseline in  sittmg/standing SBP, D B P (Instrument: H g sphygmomanometer) and FIR; 24-hour A B P M . N o t e s : Mean change in B P values extracted from text, page 20. H R data extracted from Table IV. S D of B P / H R change imputed from other trials. W D A E extracted from text, p. 21. Attempted to contact author — no response. F u n d i n g source: N o t reported. Fogari 1999 (156)  D e s i g n : R / D B / P C trial. Placebo run-in period: 4 weeks. Treatment duration: 8 weeks. Country: Italy Quality: Cochrane method = B ; Jadad score = 4 Participants: Elderly hypertensives. Sitting D B P >90 m m Fig and < 110 mm H g , and SBP > 160 mm H g for inclusion into trial. N = 54: 26 m, 28 f. Mean age: 81.8 years (range: 76 - 89). Baseline BP: 168.3/93.1 m m H g (Man. group) and 168.7/93.9 m m H g  95  (placebo group). Baseline pulse pressure: 55.2 (Man. group) and 74.8 (placebo group). Interventions: Manidipine 10 mg/day, or placebo. Primary  and secondary  sitting/standing  SBP,  outcomes: DBP  Change  (Instrument:  from  baseline in  standard  Hg  sphygmomanometer), and H R ; 24-hour A B P M (Instrument: Spacelabs 90207); E C G ; body weight. N o t e s : Sitting B P and H R data extracted from Table 2. Endpoint BPs and H R s were calculated as the weighted mean of data from week 4 and 8. S D of B P change imputed from weighted mean of S D of BPs at week 4 and 8. N o W D A E occurred. F u n d i n g source: N o t reported.  96  T a b l e 22: M i b e f r a d i l - Characteristics of i n c l u d e d studies Study  Study Description  Bernink 1996 (157)  Design: M C / R / D B / P C  trial.  Placebo run-in period: 4 weeks.  Treatment duration: 4 weeks. Country: Germany, Netherlands, Sweden, Finland, Austria, Denmark Quality: Cochrane method = B; Jadad score = 3 Participants: Sitting D B P 95-114 m m H g for inclusion into trial. N  = 202: 122 m, 80 f.  Mean age: 55.6 years (range: 1 8 - 7 0 for  inclusion). Baseline sitting D B P : 105.5 m m H g (Mib. 25 mg/day group), 104.0 mm H g (Mib. 50 mg/day group), 104.4 mm H g (Mib. 100 mg/day group), 103.9 mm H g (Mib. 150 mg/day group), 105.4 m m H g (placebo group). Baseline pulse pressure: not reported and cannot be calculated. r  Interventions: Mibefradil 25 mg/day, 50 mg/day, 100 mg/day, 150 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in sitting and standing SBP, D B P (Instrument: H g sphygmomanometer) and HR. N o t e s : Change in trough sitting D B P data and associated SDs extracted from Table 2.  Because baseline SBPs and H R s were not  reported, the corresponding data was not extracted. W D A E extracted from text, p.429. Author contacted - no additional data available. F u n d i n g source: F. Hoffman-LaRoche Ltd., Switzerland. Bursztyn 1997 (158)  Design: M C / R / D B / P C  trial.  Placebo run-in period: 4 weeks.  Treatment duration: 4 weeks. Country: Europe, Brazil, Israel Quality: Cochrane method = A ; Jadad score = 3 Participants: Sitting SBP 160-220 mm H g and D B P 90-110 m m H g for inclusion into trial.  97  Elderly patients. N = 310 (308 intention-to-treat population): 117 m, 191 f.  Mean age: 71.6 years (range: not reported; > 65 years for  inclusion). Baseline sitting B P : 176.1/99.3 m m H g (Mib. 6.25 mg/day group), 173.5/96.9 m m H g (Mib. 12.5 mg/day group), 179.3/99.7 m m H g (Mib. 25 mg/day group), 174.8/98.6 mm H g (Mib. 50 mg/day group), 176.1/100.5 m m H g (Mib. 100 mg/day group), and 174.4/99.2 mm H g (placebo group).  Baseline pulse pressure: 76.8 mm H g (Mib. 6.25  mg/day group), 76.6 mm H g (Mib. 12.5 mg/day group), 79.6 m m H g (Mib. 25 mg/day group), 76.2 m m H g (Mib. 50 mg/day group), 75.6 mm H g (Mib. 100 mg/day group), and 75.2 m m H g (placebo group). Interventions: Mibefradil 6.25 mg/day, 12.5 mg/day, 25 mg/day, 50 mg/day, 100 mg/day (50 mg/day for 1 week  100 mg/day for 3  weeks), or placebo. Primary  a n d secondary  outcomes:  Change  from  baseline in  trough/peak sitting SBP, D B P (Instrument: not reported) and H R ; E C G ; blood/urine lab tests. N o t e s : Change in trough D B P and associated SDs extracted from Table II. Change in trough SBP and associated SDs extracted from Table III. Change in H R and associated SEMs extracted from Figure 6. W D A E extracted from text, p.245. Author contacted - responded but no additional data was sent. F u n d i n g source: F. Hoffman-LaRoche Ltd. Oparil 1997 (159)  Design: M C / R / D B / P C  trial.  Placebo run-in period: 4 weeks.  Treatment duration: 4 weeks. Country: U S A Quality: Cochrane method = B; Jadad score = 3 Participants: Sitting D B P 95-110 mm H g for inclusion into trial. N = 303: 170 m, 133 f. Mean age: 51.3 years (inclusion range: 18-65). Baseline BP/pulse pressure: not reported. Interventions: Mibefradil 6.25 mg/day, 12.5 mg/day, 25 mg/day, 50  98  mg/day, 100 mg/day, 150 mg/day, 200 mg/day, or placebo. Primary  a n d secondary  outcomes:  Change  from  baseline in  trough/peak sitting SBP, D B P (Instrument: not reported) and H R ; E C G , blood/urine lab tests. N o t e s : Baseline B P s / H R s not reported, and thus, change in B P / H R data could not be used. W D A E extracted from text, p. 739. Author contacted: no data. F u n d i n g source: F. Hoffman-LaRoche Ltd., Switzerland.  99  T a b l e 23: N i c a r d i p i n e - Characteristics of i n c l u d e d studies Study  Study D e s c r i p t i o n  Asplund 1985 (160)  D e s i g n : R / D B / P C trial.  Washout period: 4 weeks.  Treatment  duration: 6 weeks. Country: Sweden Quality: Cochrane method = B; Jadad score — 4 Participants: Supine D B P 95-115 m m H g for inclusion into trial. N = 50: 36 m, 14 f. Mean age: 46 years (range: 36 - 60). Baseline BP: 161.4/113.9 mm H g (nicard. group), 154.6/112.9 m m H g (placebo group). Baseline pulse pressure: 47.5 m m H g (nicard. group), and 41.7 m m H g (placebo group). Interventions: Nicardipine 90 mg/day, or placebo. Primary  a n d secondary  supine/standing  SBP,  outcomes:  DBP  Change  (Instrument:  from  random  baseline in zero  Hg  sphygmomanometer) and H R ; lab tests; E C G . N o t e s : Change in standing B P extracted from text, p. 122S. H R s and associated SDs extracted from Fig. 2. SDs of B P / H R change imputed from weighted mean of week 4 and 6 data, extracted from Fig. 2. F u n d i n g source: N o t reported. Bellet 1987a (161)  D e s i g n : R / D B / P C trial. Washout period: 2 weeks. Placebo run-in  [Duplicate publication: Bellet 1987b (162)]  Country: France  period: 2 weeks. Treatment duration: 3 weeks.  Quality: Cochrane method = A ; Jadad score = 3 Participants: Sitting D B P 95-120 m m H g for inclusion into trial. N = 40: 27 m, 13 f. Mean age: 53 years (range: 27 - 72). Baseline standing B P : 154/107 m m H g (nicard. group), and 148/103 mm H g (placebo group). Baseline pulse pressure: 47 m m H g (nicard. group), and 45 mm H g (placebo group). Interventions: Nicardipine 100 mg/day, or placebo. Primary a n d secondary outcomes: Change from baseline in trough standing SBP, D B P , H R (Instrument: H g sphygmomanometer and  100  Sentron oscillometric device); A B P M (Instrument: Remler 2000) Notes: Change in standing B P / H R and associated SDs extracted from Table 1. Funding source: Sandoz France. De Cesaris 1993 (163)  Design: R / D B / P C trial. Washout period: 2 weeks. Placebo run-in period: 2 weeks. Treatment duration: 4 weeks. Country: Italy Quality: Cochrane method — B ; Jadad score — 3. Participants: Sitting D B P 95-116 m m H g for inclusion into trial. N = 36: 17 m, 19 f. Mean age: 55.9 years (range: 40 - 70). Baseline B P : 161.7/98.5 m m H g (nicard. group), 170.0/96.2 m m H g (placebo group). Baseline pulse pressure: 63.1 m m Fig (nicard. group), and 73.8 m m Fig (placebo group). Interventions: Nicardipine 80 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in trough supine/standing SBP, D B P (Instrument: H g sphygmomanometer); 24hr A B P M (Instrument: Spacelabs 5300). Notes: Published in Italian.  Standing BPs and associated SDs  extracted from text, p. 535.  Standing H R s and associated SDs  extracted from Fig. 2B. SD of change in B P / H R imputed from weighted mean of SD o f week4 BPs/HRs. Funding source: N o t reported. Fagan 1993 (164)  Design: M C / R / D B / P C trial. Washout period: 1 week. Placebo runin period: 2 weeks. Treatment duration: 12 weeks. Country: France Quality: Cochrane method = B ; Jadad score = 3 Participants: Supine D B P 95-114 m m H g for inclusion into trial. N =230: 127 m, 103 f. Mean age: 54.1 years (range: 22-75). Baseline B P : 157/101 m m H g (nicard. 60 mg/day group), 154/101 mm H g (nicard. 90 mg/day group), 155/101 mm H g (nicard. 120  101  mg/day group), 155/100 m m H g (placebo group).  Baseline pulse  pressure: 56 m m H g (nicard. 60 mg/day group), 53 m m H g (nicard. 90 mg/day group), 54 (nicar. 120 mg/day group), and 55 m m H g (placebo group). Interventions: Nicardipine SR 60 mg/day, 90 mg/day, 120 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in trough supine/standing  SBP,  DBP  (Instrument:  Hg  or  aneroid  sphygmomanometer) and H R ; A B P M (Instrument: SpaceLabs 90202). Notes: Change in BPs and associated SDs extracted from Fig. 1. H R data reported only for nicar. 120 mg/day.  W D A E extracted from  Table 4. Attempted to contact author — no response. Funding source: Syntex Research. Marcadet 1991 (165)  Design: R / D B / P C trial. Washout period: 2 weeks. Placebo run-in period: 15 days. Treatment duradon: 8 weeks. Country: France Quality: Cochrane method = B; Jadad score — 3 Participants: Male athletes. Sitting D B P 90-115 m m H g for inclusion into trial. N = 38: 38 m, 0 f. Mean age: 30.5 years (range: 18 — 50 for inclusion). Baseline B P : 154/98 m m Fig (nicardipine group), 155/96 m m H g (placebo group).  Baseline pulse pressure: 56 m m H g (nicardipine  group), and 59 m m H g (placebo group). Interventions: Nicardipine L P 100 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in supine SBP, D B P (Instrument: Fig sphygmomanometer) and H R at rest and on effort; maximum oxygen consumption and effort duration; weight. Notes: Published in French.  Change in BPs and associated SDs  extracted from Table III. H R data extracted from Fig. 2. Funding source: N o t reported.  102  Mazzola 1988 (166)  D e s i g n : M C / R / D B / P C trial. Washout period: duration not reported. Placebo run-in period: 1 week. Treatment duration: 6 weeks. Country: Italy Quality: Cochrane method = B; Jadad score = 4 Participants: Sitting D B P > 95 mm H g for inclusion into trial. N = 30: 19 m, 11 f Mean age: 51.8 years (range: 32 - 65). Baseline sitting BP: 158/99 m m H g (nicardipine group), 159/101 mm H g (placebo group). Baseline pulse pressure: 59 m m H g (nicardipine group), and 58 m m Fig (placebo group). Interventions: Nicardipine-SR 80 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in sitting SBP, D B P (Instrument: standard H g sphygmomanometer) H R , M B P , CI, SVI, SVR; lab parameters, E C G . N o t e s : B P / H R data extracted from Table 2.  SDs imputed from  weighted mean of week 4 and 6 data. W D A E reported. F u n d i n g source: Sandoz Prodotti Farmaceutici, Italy. Scuten 1992 (167)  D e s i g n : M C / R / D B / P C trial. Placebo run-in period: 10 days. Treatment duration: 4 weeks. Country: Italy Quality: Cochrane method — B; Jadad score = 2 Participants: Hospitalized elderly women. SBP > 160 m m H g or D B P > 95 m m H g for inclusion into trial. N = 45: 0 m, 45 f. Mean age: 76.3 years (range: not reported; > 65 years for inclusion). Baseline B P : 176.3/98.8 m m Fig (nicard. group), 166.3/95.0 m m H g (placebo group). Baseline pulse pressure: 77.5 m m H g (nicard. group), and 71.3 m m H g (placebo group). Interventions: Nicardipine (long-acting) 40 mg/day, enalapril 10 mg (with titration to 20 mg in non-responders after 1 week), or placebo. Primary a n d secondary outcomes: Change from baseline in sitting SBP, D B P (Instrument: not reported) and FIR; behavior/cognitive X  103  function tests: M M S E , G R S , S C A G . N o t e s : B P and H R data extracted from Table 1. SDs imputed from week 4 data. W D A E not reported. F u n d i n g source: Sandoz Prodotti Farmaceutici, Italy. Soro 1990 (168)  D e s i g n : R / D B / P C trial. Placebo run-in period: 2 weeks. Treatment  [Duplicate publication: De Simone 1989 (169)]  Country: Italy  duration: 8 weeks.  Quality: Cochrane method — B ; Jadad score — 3 Participants: " M i l d to moderate primary arterial hypertension" for inclusion into trial. N = 18: 8 m, 10 f. Mean age: years (range: 27 - 65). Baseline standing B P : 164/106 m m H g (Nicardipine group), and 167/105 m m Fig (placebo group). Baseline pulse pressure: 58 mm Fig (Nicardipine group), and 62 mm H g (placebo group). Interventions: Nicardipine 60 mg/day, or placebo. Primary  and secondary  outcomes:  Change  from baseline in  supine/standing SBP, D B P and H R (Instrument: automatic device: Sentron); body weight; serum electrolytes, renin, aldosterone, P T H , cholesterol, fasting blood glucose, L V function by echocardiography. N o t e s : Standing office BPs and SDs extracted from text, p. 135 of Soro f 990. SDs of B P change imputed from week 8 data. H R s and W D A E s extracted from De Simone 1989. Author (De Simone) contacted — clarified duplicate publication. F u n d i n g source: N o t reported.  104  Table 24: Nifedipine - Characteristics of included studies Study  Study Description  Carr 1992a (170)  Design: 2-centre R / D B / P C  [Duplicate publication: Carr 1992b (171)]  trial.  Placebo run-in: 3-6 weeks.  Treatment duration: 6 weeks. Country: U S A Quality: Cochrane method = B ; Jadad score = 3 Participants: sitting D B P 95-119 m m H g for inclusion into trial. N = 207: 79 m, 128 f. Mean age: 46.2 years (range: not reported). Baseline BP: 155.4/103.7 mm H g (Nif. 20 mg/day group), 153.0/104.0 mm  H g (Nif. 50 mg/day group), 156.8/104.9 m m H g (Nif. 100  mg/day group), and 153.2/104.1 m m H g (placebo group). Baseline pulse pressure: 51.7 m m H g (Nif. 20 mg/day group), 49 m m H g (Nif. 50 mg/day group), 51.9 m m H g (Nif. 100 mg/day group), and 49.1 mm H g (placebo group). Interventions: Nifedipine SR 20, 50, and 100 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in trough sitting, standing, supine SBP, D B P (Instrument: not reported) and H R ; A B P M (Instrument: D e l Mar Avionics Pressurometer IV). Notes: Change in supine B P data extracted from Table II. S D of B P change imputed from other trials. W D A E extracted from Table IV. Author contacted - no data. Funding source: N o t reported. De Simone 1984 (172)  Design: R / D B / P C  [Duplicate publication: De Simone 1985 (173)]  Country: Italy  trial.  Washout period: 2 weeks.  Treatment  duration: 8 weeks.  Quality: Cochrane method — B ; Jadad score = 2 Participants: D B P 94-114 m m H g for inclusion into trial. N = 20: 11 m, 9 f. Mean age: 47 years (range: 35 - 60). Baseline standing B P : 156.1/109 m m H g (Nif. group), 148.7/107.2 mm H g (placebo group). Baseline pulse pressure: 54.3 m m H g (Nif. group), and 41.5 mm H g (placebo group).  105  Interventions: Nifedipine SR 40 mg/day, or placebo. Primary  a n d secondary  outcomes:  Change  from  baseline in  supine/standing SBP, D B P (Instrument: not reported); left ventricular mass, cross sectional area, end-systolic and end-diastolic volumes, ejection fraction and mean velocity of circumferential fiber shortening as measured by echocardiogram. N o t e s : Standing B P and H R data extracted from Table 1.  SDs of  B P / H R change imputed from week 8 data. F u n d i n g source: N o t reported. Eggertsen 1982 (174)  Design: R / D B / P C  trial.  Placebo run-in: 4 weeks.  Treatment  duration: 8 weeks, followed by 12 weeks of added metoprolol. Country: Sweden Quality: Cochrane metiiod = B; Jadad score = 3 Participants: "Essential hypertension". N = 26 (m:f ratio not reported). Mean age: 51.3 years (range: 28 - 60). Baseline B P : 157/106 mm H g (Nif. group), and 155/104 m m H g (placebo group). Baseline pulse pressure: 51 mm H g (Nif. group), and 51 m m H g (placebo group). Interventions: Nifedipine 30 mg/day, or placebo. Primary  and secondary  outcomes:  Change  from  baseline in  supine/standing SBP, D B P (Instrument: H g manometer) and H R . N o t e s : Supine BPs at week 8 extracted from Table 1. change imputed from other trials.  S D of B P  H R data not reported.  WDAE  extracted from text, p. 390. Author contacted - no additional data available. Number of patients in each group was obtained from a published meta-analysis (93). F u n d i n g source: N o t reported. Fadayomi 1986 (175)  D e s i g n : R / D B / P C trial. Washout period: 2 weeks. Placebo run-in: 2 weeks. Treatment duration: 6 weeks. Country: Nigeria Quality: Cochrane method = B; Jadad score = 3  106  Participants:  supine  DBP >  110 m m H g (newly diagnosed  hypertensives) or > 100 m m Fig (diagnosed 3-6 months earlier and inadequately controlled) for inclusion into trial. N = 32: 18 m, 14 f. Mean age: 48 years (range: 37 - 59). Baseline B P : 181.3/114.7 m m H g (Nif. group), and 179.5/114.0 m m H g (placebo group).  Baseline pulse pressure: 66.6 m m H g (Nif.  group), and 65.5 m m H g (placebo group). Interventions: Nifedipine (Adalat Retard) 40 mg/day, or placebo. Primary  and  secondary  supine/standing  SBP,  outcomes: DBP  Change  from  (Instrument:  baseline in  standard  Hg  sphygmomanometer) and H R ; body weight. Notes: Individual patient B P data for 30 patients extracted from Table 1 and 2. S D o f B P change calculated from this data.  H R data not  extracted, as data was reported for nifedipine group only.  WDAE  extracted from text, p. 468. Author contacted - replied but no missing data was supplied. Funding source: Bayer Pharmaceuticals Ltd., Nigeria. Feig 1993 (176)  Design: M C / R / D B / P C trial.  Placebo run-in: 4 weeks.  Treatment  duration: 6 weeks.  Country: U S A Quality: Cochrane method = B; Jadad score = 4 Participants: supine D B P 95-114 m m Fig for inclusion into trial. N = 233: 130 m, 103 f. Mean age: 55 years (range: 19 - 75). Baseline standing B P : 150/100 m m Fig (Nif. 30 mg/day group), 150/100 m m H g (Nif. 60 mg/day group), 153/101 m m H g (Nif. 90 mg/day group), and 149/100 m m Fig (placebo group). Baseline pulse pressure: 50 m m H g (Nif. 30 mg/day group), 50 m m H g (Nif. 60 mg/day group), 52 m m H g (Nif. 90 mg/day group), and 49 m m H g (placebo group). Interventions: Nifedipine coat-core 30 mg/day, 60 mg/day, 90 mg/day, or placebo.  107  Primary  and secondary  outcomes:  Change  from  baseline in  supine/standing trough SBP, D B P (Instrument: not reported), and HR. N o t e s : Change in standing B P data extracted from Table II. Author contacted: SEs of B P change obtained. F u n d i n g source: Miles Inc., U S A . Ferrera 1984 (177)  D e s i g n : R / D B / P C trial. Washout period: 2 weeks. Placebo run-in: 2 weeks. Treatment duration: 8 weeks. Country: Italy Quality: Cochrane method = B; Jadad score = 2. Participants: Primary uncomplicated arterial hypertension of mild to moderate degree for inclusion into trial. N = 40: 51 m, 33 f. Mean age: 47 years (range: not reported). Baseline B P : 164/108 m m Fig (Nif. group), and 151/104 mm H g (placebo group). Baseline pulse pressure: 56 m m H g (Nif. group), and 47 mm Ffg (placebo group). Interventions: Nifedipine (slow release) 20 mg/day, or placebo. Primary  a n d secondary  supine/standing  DBP,  outcomes: SBP  Change  from  (Instrument:  baseline in random-zero  sphygmomanometer) and H R ; left ventricular mass, ejection fraction, end diastolic volume, end systolic volume, mean rate of circumferential fiber shortening by echocardiogram; systolic time intervals. N o t e s : BPs, H R s and associated SDs extracted from Table 1. S D of B P / H R change imputed from week 8 data. Attempted to contact author — no response.  Number of patients in  each group was obtained from a published meta-analysis (93). F u n d i n g source: N o t reported. Harder 1994 (178)  D e s i g n : M C / R / D B / P C trial. Washout period: 2 weeks. Placebo runin: 2 weeks. Treatment duration: 8 weeks. Country: Germany Quality: Cochrane method = B; Jadad score = 3  108  Participants: D B P 95-115 m m H g for inclusion into trial. N — 88: 51 m, 33 f. Mean age: 54 years (range: not reported). Baseline B P : 155/102 m m H g (Nif. group), and 155/102 m m H g (placebo group). Baseline pulse pressure: 53 m m H g (Nif. group), and 53 m m H g (placebo group). Interventions: Nifedipine (slow release) 60 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in sitting trough  D B P , SBP (Instrument:  H g sphygmomanometer;  semi-  automatic auscultatory device - Tonoprint) and H R . N o t e s : BPs (manual readings) and H R s extracted from Table I. S D of B P / H R change imputed from week 8 data.  W D A E extracted from  text, p. 135. F u n d i n g source: N o t reported. Jueng 1987 (179)  D e s i g n : M C / R / D B / P C trial. Washout period: 1 week. Placebo runin: 3 weeks. Treatment duration: 2-week dose titration, followed by 8 weeks. Country: U S A Quality: Cochrane method = B; Jadad score = 4 Participants: sitting D B P 95-110 m m H g for inclusion into trial. N = 29: 22 m, 7 f. Mean age: 52.4 years (range: 27 - 83). Baseline B P : 159/102 mm H g (nifed. group), 153/105 m m H g (placebo group). Baseline pulse pressure: 57 mm Fig (nifed. group), and 48 mm H g (placebo group). Interventions: Nifedipine G I T S 30 mg/day (2 weeks)-^ 60 mg/day, hydrochlorothiazide 25 mg/day (2 weeks)  50 mg/day, or placebo.  Primary and secondary outcomes: Change from baseline in sitting trough  SBP, D B P (Instrument:  not  reported),  H R , metabolic  parameters. N o t e s : B P s / H R s extracted from Table 2. SDs of change in B P s / H R s imputed from week 8 values. W D A E extracted from text, p. 697. F u n d i n g source: Pfizer Pharmaceuticals, U S A .  109  Serradimigni 1985 (180)  D e s i g n : M C / R / D B / P C trial.  Placebo run-in: 4 weeks.  Treatment  duration: 24 weeks. Country: France Quality: Cochrane method = B; Jadad score = 3 Participants: sitting D B P 95-115 m m H g for inclusion into trial. N = 177: 95 m, 82 f. Mean age: 55.6 years (range: 38 - 70). Baseline B P : 178.1/103.6 mm H g (Nif. group), 177.2/104.5 m m H g (placebo group). Baseline pulse pressure: 74.5 m m Ffg (Nif. group), and 72.7 mm H g (placebo group). Interventions: Nifedipine 40 g/day, Acebutolol 200 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in supine SBP, D B P (Instrument: Fig sphygmomanometer), H R , body weight. N o t e s : Published in French. D B P , SBP and H R at baseline and week 4, 8, 12 extracted from Fig. 5, 6, and 7, respectively. SDs of B P / F I R change imputed from other trials. W D A E extracted from Table 2. F u n d i n g source: N o t reported.  Toal 1997 (181)  D e s i g n : M C / R / D B / P C trial. Placebo run-in: 2-3 weeks. Treatment duration: 4 weeks. Country: Canada Quality: Cochrane method = B; Jadad score = 2 Participants: sitting D B P 95-114 mm Fig for inclusion into trial. N - 187: 118 m, 69 f. Mean age: 55 years (> 18 for inclusion). Baseline BP: 149.1/98.4 mm Fig (Nif. group), and 148.0/98.7 m m H g (placebo group). Baseline pulse pressure: 50.7 m m H g (Nif. group), and 49.3 mm Fig (placebo group). Interventions: Nifedipine GfTS 20 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in sitting trough SBP, D B P (Instrument: not reported) and H R ; A B P M (Instrument: Spacelabs 90207). N o t e s : B P and H R data extracted from Table 2.  SDs of B P / H R  110  change imputed from SDs of week 4 B P s / H R s . Funding source: Bayer Inc., Canada. Zachariah 1990 (182)  Design: M C / R / D B / P C trial.  Placebo run-in: 3 weeks.  Treatment  duration: 8 weeks.  Country: U S A Quality: Cochrane method = B; Jadad score = 4 Participants: sitting D B P 95-104 m m H g for inclusion into trial. N = 29: 19 m, 10 f. Mean age: 47.3 years (range: not reported). Baseline B P : 137/98 m m H g (nifed. 30 mg/day group), 141/98 m m Hg  (nifed. 60 mg/day group), 133/97 m m H g (placebo group).  Baseline pulse pressure: 39 m m H g (nifed. 30 mg/day group), 43 (nifed. 60 mg/day group), and 36 m m H g (placebo group). Interventions: Nifedipine (sustained release) 30 mg/day, 60 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in sitting trough SBP, D B P (Instrument: H g sphygmomanometer), 24-hr A B P M (Pressurometer III, D e l Mar Avionics). Notes: BPs extracted from Table II. S D of change in B P imputed from week 8 BPs. W D A E extracted from text, p. 1016. Funding source: N o t reported. Zanchetti 1993 (183) [Multiple publications: Zanchetti 1994a (184) ; Zanchetti 1994b (185) , Zanchetti 1994c (186)]  Design: M C / R / D B / P C  trial.  Placebo run-in period: 2 weeks.  Treatment duration: 4 weeks. Country: Italy Quality: Cochrane method = B; Jadad score = 2 Participants: sitting D B P 95-114 m m H g for inclusion into trial. N = 126: 65 m, 61 f. Mean age: 56.5 years (range: 25 - 74). Baseline B P : 154.3/96.6 m m H g (Nif. 30 mg/day group), 150.1/94.5 mm H g (Nif. 60 mg/day group), 146.0/90.1 m m H g (placebo group). Baseline pulse pressure: 57.7 mm H g (Nif. 30 mg/day group), 55.6 m m H g (Nif. 60 mg/day group), and 55.9 m m H g (placebo group). Interventions: Nifedipine GITS 30 mg/day, 60 mg/day, or placebo.  Primary and secondary outcomes: Change from baseline in sitting trough SBP, D B P (Instrument: H g sphygmomanometer) and H R ; A B P M (Instrument: Spacelabs 90202 or 90207). N o t e s : Change in clinic B P and associated SEMs extracted from text, p. 334 of Zanchetti 1993.  W D A E extracted from text, p. 54 of  Zanchetti 1994a. F u n d i n g source: N o t reported.  112  T a b l e 25: N i l v a d i p i n e - Characteristics of i n c l u d e d studies Study Hoffmann 1997 (187)  Study Description Design: M C / R / D B / P C  trial. Placebo run-in period: 3-5 weeks.  Treatment duration: 8 weeks. Country: Germany Quality: Cochrane method = B; Jadad score — 3 Participants: Sitting D B P 95-115 m m H g for inclusion into trial. N = 168 (ITT): 79 m, 89 f. Mean age: 56.6 years (range: 25 - 80). Baseline BP: 166.8/101.9 m m H g (Nilv. 8 mg/day group), 167.1/101.7 mm H g (Nilv. 16 mg/day group), 164.2/101.3 m m H g (placebo group). Baseline pulse pressure: 64.9 m m H g (Nilv. 8 mg/day group), 65.4 mm Pig (Nilv. 16 mg/day group), and 62.9 m m H g (placebo group). Interventions: Nilvadipine 8 mg/day, 16 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in sitting and standing SBP, D B P (Instrument: H g sphygmomanometer), and H R ; A B P M (Instrument: SpaceLabs, model 90207.32). N o t e s : Trough changes in sitting B P and associated S E M s extracted from Table 3. W D A E extracted from Table 4. F u n d i n g source: N o t reported.  Weir 1990 (188)  Design: M C / R / D B / P C  trial. Placebo run-in period: 2-4 weeks.  Treatment duration: 4 weeks. Country: U S A Quality: Cochrane method = A ; Jadad score = 3 Participants: Sitting D B P 100-115 m m H g for inclusion into trial. N = 84: 65 m, 19 f. Mean age: 51.0 years (range: 28 - 70). Baseline sitting B P : 160.6/103.0 mm H g (Nilv. 18 mg/day group), 152.8/103.9 mm H g (Nilv. 24 mg/day group), 154.3/103.4 mm H g (Nilv. 30 mg/day group), and 156.0/103.9 m m H g (placebo group). Baseline pulse pressure: 57.6 mm H g (Nilv. 18 mg/day group), 48.9 mm H g (Nilv. 24 mg/day group), 50.9 m m Fig (Nilv..30 mg/day), and  113  52.1 m m H g (placebo group). Interventions:  Nilvadipine 18 mg/day, 24 mg/day, 30 mg/day, or  placebo. Primary a n d secondary outcomes: Change from baseline in sitting SBP, D B P (fnstrument: H g sphygmomanometer), M A P , and H R ; E C G ; blood and urine lab tests. N o t e s : Changes in sitting D B P and associated SEMs from week 3 and 4 extracted from Table I V and weighted means were calculated. Changes in S B P / H R not reported.  W D A E in each group not  reported. F u n d i n g source: N o t reported.  114  T a b l e 26: N i s o l d i p i n e - Characteristics of included studies Study Opie 1997 (ANCHOR) (189)  Study D e s c r i p t i o n Design: M C / R / D B / P C  trial.  Placebo run-in period: 4 weeks.  Treatment duration: 6 weeks. Country: South Africa Quality: Cochrane method = B; Jadad score — 3 Participants: Supine D B P 95-114 m m H g for inclusion into trial. N = 206: 92 m, 114 f (efficacy analysis). Mean age: 52.1 years (range: 20 - 75). Baseline standing B P : 159.7/107.9 mm H g (Nisol. 10 mg/day group), 163.8/107.4 mm H g (Nisol. 20 mg/day group), 161.7/107.4 mm H g (Nisol. 30 mg/day group), and 160.5/105.1 m m H g (placebo group). Baseline pulse pressure: 51.8 mm H g (Nisol. 10 mg/day group), 56.4 mm H g (Nisol. 20 mg/day group), 54.3 mm H g (Nisol. 30 mg/day group), and 55.4 mm H g (placebo group). Interventions:  Nisoldipine coat-core tablet 10 mg/day, 20 mg/day,  30 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in trough supine/standing SBP, D B P (Instrument: automatic device —Dinamap), and supine H R ; A B P M (Instrument: SpaceLabs 90202). N o t e s : Changes in standing B P and supine H R data extracted from Table 2. S D of changes in B P / H R imputed by calculating pooled S D from reported 95% C I of difference between treatment and placebo. Author contacted — responded by referring to statistics company (no response). F u n d i n g source: N o t reported.  115  T a b l e 27: N i t r e n d i p i n e - Characteristics of included studies Study Asmar 1992 (190)  Study D e s c r i p t i o n D e s i g n : R / D B / P C trial. Placebo run-in period: 15 days. Treatment  [Duplicate publication: Asmar 1993 (191)]  Country: France  duration: 4 weeks.  Quality: Cochrane method = B; Jadad score = 2 Participants: Supine D B P > 95 m m Fig for inclusion into trial. N- = 17: 14 m, 3 f. Mean age: 50 years (range: 39 - 64). Baseline B P : 160/103 m m H g (Nitren. group), and 158/104 m m H g (placebo group). Baseline pulse pressure: 57 mm H g (Nitren. group), and 54 m m H g (placebo group). Interventions: Nitrendipine 20 mg/day, or placebo. Primary a n d secondary outcomes: Change from baseline in trough supine/standing SBP, D B P (Instrument: Fig sphygmomanometer), and FIR, A B P M (Instrument: Novacor model Diasys 200-R), arterial pulse wave velocity. N o t e s : Change in B P / H R and associated SDs extracted from Table 1. W D A E not reported. Audior  contacted:  duplicate publication clarified;  no  access  to  additional data. F u n d i n g source: I N S E R M & Ministere de la Recherche, France. Ferrera 1985 (192)  D e s i g n : R / D B / P C trial. Washout period: 2 weeks. Placebo run-in period: 2 weeks. Treatment duration: 8 weeks. Country: Italy Quality: Cochrane method = B; Jadad score = 2 Participants: "hypertension" with SBP <200 m m H g and D B P <120 mm Fig for inclusion into trial. N = 30: 20 m, 10 f. Mean age: 47.3 years (range: 24 - 62). Baseline standing B P : 159/113 m m H g (Nitren. group), and 149/107 m m H g (placebo group). Baseline pulse pressure: 46 m m H g (Nitren. group), and 42 m m Fig (placebo group).  116  Interventions: Nitrendipine 20 mg/day, or placebo. Primary a n d secondary outcomes: Change from baseline in trough supine/standing SBP, D B P (Instrument: automatic recorder), and H R , exercise testing; L V mass, systolic function and ejection fraction by echocardiography. N o t e s : Standing B P / F I R data and associated SDs extracted from Table f and text, p. 436. SDs of B P change imputed from SDs of week 8 BPs. W D A E extracted from text, p. 437. F u n d i n g source: N o t reported. Fodor1991 (193)  D e s i g n : R / D B / P C trial. Washout period: 2 weeks. Placebo run-in period: 2 weeks.  Treatment duration: 4 weeks, followed by dose  doubling in non-responders for 6 weeks. Country: Canada Quality: Cochrane method = B; Jadad score = 3 Participants: Supine D B P 95-104 mm H g for inclusion into trial. N = 80: 44 m, 36 f. Mean age: 52.8 years (range: 21 - 70). Baseline  supine  B P : 155.7/98.2  mm  H g (Nitren. group),  and  155.0/97.4 m m H g (placebo group). Baseline pulse pressure: 57.5 m m Fig (Nitren. group), and 57.6 m m H g (placebo group). Interventions: Nitrendipine 10 mg/day, or placebo. Primary  a n d secondary  supine/standing  outcomes:  trough  SBP,  Change DBP  from  baseline in  (Instrument:  Hg  sphygmomanometer), and FIR. N o t e s : Baseline BP/FIRs and associated SEMs extracted from Table 1.  Week 4 supine BPs/FIRs and associated SEMs extracted from  Figure 3.  SDs of B P change imputed from SDs of week 4 BPs.  W D A E extracted from text, p. 226. F u n d i n g source: N o t reported. Gerntsen 1998 (194)  D e s i g n : R / D B / P C trial. Washout period: > 3 weeks. Placebo run-in period: 4 weeks.  Treatment duration: 4 weeks, followed by dose  doubling in non-responders for 44 weeks.  117  Country: Netherlands Quality: Cochrane method = B; Jadad score = 4 Participants: Non-insulin-dependent diabetes mellitus patients. D B P 90-115 mm H g and SBP < 200 m m H g for inclusion into trial. N = 81 (excludes patients in enalapril group): 47 m, 34 f. Mean age: 64.3 years (range: not reported). Baseline standing BP: 168/90 m m Pig (Nitren. group), and 166/93 mm Pig (placebo group).  Baseline pulse pressure: 78 m m H g (Nitren.  group), and 73 mm Pig (placebo group). Interventions: Nitrendipine 20 mg/day, enalapril 10 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in trough supine/standing SBP, D B P (fnstrument: automatic device - Dinamap), and H R , fasting blood glucose, serum lipids, echocardiogram. Notes: BPs and associated SDs extracted from Table 1 (baselines) and Figure 1 (week 4 data). S D of B P change imputed from S D of week 4 BPs. Funding source: Bayer, Nederland.  118  Gregorio 1991 (195)  D e s i g n : R / D B / P C trial.  Washout period: 2 weeks.  Treatment  duration: 12 weeks. Country: Italy Quality: Cochrane method = B; Jadad score = 2 Participants: Type 2 diabetes mellitus patients. D B P 95-115 mm H g for inclusion into trial. N = 30: 18 m, 12 f. Mean age: not reported (range: 52 — 74 years). Baseline standing B P : 166.1/112.8 m m H g (Nitren. group), and 158.9/113.5 m m H g (placebo group).  Baseline pulse pressure: 53.3  m m H g (Nitren. group), and 45.4 m m H g (placebo group). Interventions: Nitrendipine 20 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in trough supine/standing SBP, D B P (Instrument: H g sphygmomanometer), and HR;  lipid  profile,  glycemic homeostatis  and  other  metabolic  parameters. N o t e s : Published in Italian. BPs and associated SDs extracted from Fig. 3. S D of B P change imputed from weighted mean S D of BPs at weeks 3, 4, 6, 8,10,12. Attempted to contact author — no reply. F u n d i n g source: N o t reported. Kuschnir 1988 (196)  D e s i g n : R / D B / P C crossover trial. Placebo run-in period: 1-3 weeks. Treatment duration: 3 weeks. Country: Argentina Quality: Cochrane method = B; Jadad score = 2 Participants: M i l d to moderate  ( W H O stage I or II) essential  hypertension for inclusion into trial. N = 20: 14 m, 3 f. Mean age: 48.9 years (range: 33 - 60). Baseline standing BP: 167/106 m m H g (Nitren. group), and 168/107 m m H g (placebo group). Baseline pulse pressure: 61 m m Fig (Nitren. group), and 61 mm H g (placebo group). Interventions: Nitrendipine 20 mg/day, or placebo.  119  Primary and secondary outcomes: Change from baseline in trough supine/standing SBP, D B P (Instrument: H g sphygmomanometer), and H R , peripheral hemodynamic parameters by radionuclide techniques. Notes: B P s / H R s and associated SDs extracted from Table 2. SDs o f B P / H R change imputed from other trials. Funding source: N o t reported. Lederle 1991 (197) [Duplicate publication: Klaus 1990 (198)]  Design:  MC/R/DB/PC  trial.  Placebo run-in period: 2 weeks.  Treatment duration: 6 weeks. Country: Germany Quality: Cochrane method = B; Jadad score = 3 Participants: W F I O stage f mild essential hypertension for inclusion into trial. N = 141: 88 m, 53 f. Mean age: 54.8 years (range: not reported). Baseline sitting BP: 161.4/99.6 m m H g (Nitren. group), and 161.3/99.6 mm H g (placebo group). Baseline pulse pressure: 61.8 m m H g (Nitren. group), and 61.7 m m Fig (placebo group). Interventions: Nitrendipine 10 mg/day, or placebo. Primary and secondary outcomes: Number of treatment responders (DBP reduced by > 10 m m H g and/or to a value o f < 90 m m Hg) after 6 weeks o f treatment.  Change from baseline in sitting trough  SBP, D B P (Instrument: sphygmomanometer), and H R . Notes: B P s / H R s and associated SDs extracted from text, p. S49 and figure 1. SDs of S B P / H R change imputed from weighted mean of week 4 and 6 SBPs/FIRs. W D A E extracted from text, p. S50. Funding source: N o t reported. Maclean 1990 (199)  Design: M C / R / D B / P C trial. Washout period: 2-4 weeks. Placebo run-in period: 3 weeks. Treatment duration: 6 weeks, followed by dose doubling in nonresponders for 6 weeks.  Country: U K Quality: Cochrane method = B; Jadad score = 3 Participants: Sitting D B P 95-115 m m H g for inclusion into trial.  120  N  — 64 (excluding atenolol and combination groups): 29 m, 35 f.  Mean age: 51 years (range: 18-75). Baseline sitting BP: 171/108 m m H g (Nitren. group), and 171/108 mm H g (placebo group).  Baseline pulse pressure: 63 m m H g (Nitren.  group), and 63 mm H g (placebo group). Interventions: Nitrendipine 20 mg/day, atenolol 50 mg/day, atenolol 50 mg + nitrendipine 20 mg combination, or placebo. Primary a n d secondary outcomes: Change from baseline in trough sitting/standing SBP, D B P (Instrument:  Hawkesley random-zero  sphygmomanometer), and H R . N o t e s : Sitting B P s / H R s extracted from Table 4. SDs of baseline and week 6 BPs extracted from Figure 2. S D of B P change imputed from week 6 S D of BPs. (95% C I of differences between treatment and placebo were reported but mean differences were not reported). W D A E extracted from text, p. 457. F u n d i n g source: N o t reported. PaoHsso 1991 (200)  D e s i g n : R / D B / P C trial.  Washout period: > 3 weeks.  Treatment  duration: 60 days. Country: Italy Quality: Cochrane method — B ; Jadad score = 2 Participants: Elderly patients with I S H . N = 20: 11 m, 9 f. Mean age: 76.8 years (range: not reported). Baseline B P : 180/92 m m H g (Nitren. group), and 178/91 mm H g (placebo group). Baseline pulse pressure: 88 m m Fig (Nitren. group), and 87 m m Fig (placebo group). Interventions: Nitrendipine 20 mg/day, or placebo. Primary a n d secondary outcomes: Change from baseline in SBP, DBP  (Instrument:  not  reported),  and  H R ; hemodynamic  and  laboratory tests; oral glucose tolerance test. N o t e s : B P s / H R s and associated SEMs extracted from table, p. 696. SDs of B P change imputed from SDs of week 8 BPs.  121  Funding source: N o t reported. Roca-Cusachs 2001 (201)  Design: M C / R / D B / P C  factorial-design trial.  Washout period: 1  week. Placebo run-in period: 2 weeks. Treatment duration: 6 weeks. Country: Spain Quality: Cochrane method = B; Jadad score = 2 Participants: Sitting D B P 90-109 mm H g for inclusion into trial. N = 342 (per-protocol): 137 m, 205 f. Mean age: 55.6 years (range for inclusion: 18-70). Baseline BP: 158.3/98.6 m m H g (all groups). Baseline pulse pressure: 59.7 m m H g (all groups). Interventions: Nitrendipine 5 mg/day, 10 mg/day, 20 mg/day, Enalapril  5  mg/day,  10  mg/day,  20  mg/day,  all 9  possible  combinations thereof, or placebo. Primary and secondary outcomes: Change from baseline in sitting trough D B P , SBP (Instrument: Fig sphygmomanometer), and FIR; rate of responders (< 140/90 m m H g or reduction of >20/10 m m Fig from baseline). Notes: Change in BPs extracted from figure 1. SDs of SBP change imputed from S D of baseline SBP. SDs of D B P change imputed from other trials. W D A E for each group not reported. Funding source: Vita-Invest S.A., Spain.  122  T a b l e 28: Pranidipine - Characteristics of i n c l u d e d studies Study Rosenthal 1996 (202)  Study Description D e s i g n : M C / R / D B / P C trial.  Placebo run-in: 4 weeks.  Treatment  duration: 4 weeks. Country: Germany Quality: Cochrane method =B; Jadad score = 4 Participants: Sitting D B P 95-114 mm H g for inclusion into trial N = 176: 97 m, 79 f. Mean age: 52.0 years (inclusion range: 20 - 70). Baseline B P : 159/103 m m H g (Pran. 1 mg/day group), 154/102 m m H g (Pran. 2 mg/day group), 156/103 mm H g (Pran. 4 mg/day group), 151/103 mm H g (Pran. 8 mg/day group), 153/103 m m H g (placebo group). Baseline pulse pressure: 56 mm H g (Pran. 1 mg/day group), 52 mm H g (Pran. 2 mg/day group), 53 mm H g (Pran. 4 mg/day group), 48 m m Pig (Pran. 8 mg/day group), and 50 mm H g (placebo group). Interventions: Pranidipine 1 mg/day, 2 mg/day, 4 mg/day, 8 mg/day, or placebo. Primary a n d secondary outcomes: Change from baseline in sitting trough  SBP, D B P and  PIR (Instrument:  automated  device —  Tonoprint). N o t e s : B P data extracted from text, page 61 and 64. W D A E extracted from Table 6. S D of B P change imputed from SDs of BPs at week 4. F u n d i n g source: N o t reported.  123  Table 29: Tiapamil - Characteristics of included studies Study Blanchett 1991 (203)  Study Description Design: M C / R / D B / P C trial.  Placebo run-in: 4 weeks.  Treatment  duration: 6 weeks, followed by 1 week washout. Country: U S A Quality: Cochrane method =B; Jadad score = 3 Participants: Sitting D B P 95-115 m m H g for inclusion into trial N = 58: 36 m, 22 f. Mean age: 50.5 years (range: 24 - 67). Baseline BP: 153/103 m m H g (Level. I group), 150/100 m m H g (Level II group), 155/104 m m Fig (placebo group). Baseline pulse pressure: 50 mm H g (Level I group), 50 mm H g (Level II group), and 51 mm H g (placebo group). Interventions: Tiapamil 300 mg/day or 600 mg/day ("Level I"), 900 mg/day or 1200 mg/day ("Level II"), or placebo. Primary  and  secondary  standing/sitting/supine  outcomes: Change  from  baseline in  trough SBP, D B P and FIR  (Instrument:  sphygmomanometer, A B P M (Instrument: Spacelabs 5200) Notes: BPs, H R s and associated SDs extracted from table 2 (reported in dosage groups 300-600 mg/day and 900-1200 mg/day).  WDAE  extracted from text, p. 61. S D of change in B P s / H R s imputed from week 6 BPs. Author contacted to obtain # of patients on each dose — no response. Funding source: Hoffman La-Roche, Inc., U S A .  124  T a b l e 30: V e r a p a m i l - Characteristics of i n c l u d e d studies Study Carr 1991 (204)  Study Description D e s i g n : M C / R / D B / P C trial.  Placebo run-in period: 3-6 weeks.  Treatment duration: 4 weeks. Country: U S A Quality: Cochrane method = B; Jadad score = 2 Participants: Supine D B P 95-119 mm H g for inclusion into trial N = 107: m, f. Mean age: 46 years (range: not reported). Baseline supine B P : 152.9/103.3 mm H g (ver. 120 mg/day group), 157.9/103.3 m m H g (ver. 240 mg/day group), 148.3/100.7 mm Pig (ver. 480 mg/day group), and 156.9/103.2 m m H g (placebo group). Baseline pulse pressure: 49.6 m m Pig (ver. 120 mg/day group), 54.6 mm H g (ver. 240 mg/day group), 47.6 mm H g (ver. 480 mg/day group), and 53.7 mm H g (placebo group). Interventions: Verapamil Q D 120 mg/day, 240 mg/day, 480 mg/day, or placebo. Primary  and secondary  supine/sitting/standing sphygmomanometer)  outcomes: SBP,  and  Change  DBP  PIR; A B P M  from  baseline in  (Instrument: (Instrument:  Hg  D e l Mar  Pressurometer III), serum biochemical markers. N o t e s : Change in supine B P , PIR and associated SDs extracted from Table f. F u n d i n g source: N o t reported. D e Quattro 1997a (205) [Multiple publications: D e Quattro 1997b (206) ;Levine 1997 (207) ]  D e s i g n : M C / R / D B / P C 4x4 factorial trial. Placebo run-in period: 4 weeks. Treatment duration: 6 weeks. Country: U S A Quality: Cochrane method = A ; Jadad score = 3 Participants: Sitting D B P 95-114 m m H g for inclusion into trial. N = 726: 456 m, 270 f. Mean age: 54.7 years (range: 24 - 83). Baseline sitting B P : 148.4/100.1 m m H g (verap. 120 mg/day group), 153.4/99.3 m m H g (verap. 180 mg/day group), 150.3/100.7 mm H g  125  (verap 240 mg/day group), 153.7/100.3 m m H g (placebo group). Baseline pulse pressure: 48.3 m m H g (verap. 120 mg/day group), 54.1 (verap 180 mg/day group), 49.6 m m H g (verap. 240 mg/day group), and 53.4 m m H g (placebo group). Interventions: Verapamil SR 120 mg/day, 180 mg/day, 240 mg/day, Trandolapril 0.5 mg/day, 2 mg/day, 8 mg/day, all combinations thereof, or placebo. Primary and secondary outcomes: Change from baseline in sitting trough SBP, D B P (Instrument: standard H g sphygmomanometer), H R Notes: BPs extracted from Table III (DeQuattro 1997a). SDs o f SBP change imputed from supine baseline SBPs.  S D of D B P change  imputed from other trials. Funding source: K n o l l Pharmaceutical Co., U S A . Levine 1995 (208)  Design: M C / R / D B / P C 3x2 factorial trial. Placebo run-in period: 4-5 weeks. Treatment duration: 4 weeks.  Country: USA Quality: Cochrane method = B; Jadad score — 4 Participants: Sitting D B P 95-114 m m H g for inclusion into trial. N = 186: 105 m, 81 f. Mean age: 52.7 years (range: 24 - 80). Baseline  B P : 150.4/101.2  m m H g (verap.120  mg/day  group),  151.1/100.4 m m H g (verap. 240 mg/day), 149.8/100.2 m m H g (placebo group).  Baseline pulse pressure: 49.2 m m H g (verap. 120  mg/day group), 50.6 m m H g (verap. 240 mg/day group), and 49.6 m m H g (placebo group). Interventions:  Verapamil 120 mg/day, 240 mg/day, enalapril 10  mg/day, verapamil + enalapril 120/10 mg/day, 240/10 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in trough sitting SBP, D B P (Instrument: standard H g sphygmomanometer), E C G , lab tests, quality of life questionnaire. Notes: Change in SBP and D B P with associated SDs extracted from  126  Table 2 and 3, respectively. W D A E extracted from text, p. 497. F u n d i n g source: Lederle Laboratories, U S A . McMahon 1989 (209)  D e s i g n : M C / R / D B / P C trial.  Placebo run-in period: 1-4 weeks.  Treatment duration: 6 weeks. Country: U S A Quality: Cochrane method = B; Jadad score = 3 Participants: Supine D B P 95-114 mm H g for inclusion into trial. N = 221: 127 m, 94 f. Mean age: 54.8 years (>f 8 for inclusion). Baseline supine B P : 159.6/101.8 m m Fig (verap. 60 mg/day group), 152.6/99.8 m m H g (verap. 120 mg/day group), 157.7/101.4 m m H g (verap. 240 mg/day group), 151.8/99.6 mm H g (verap. 480 mg/day), 156.2/100.7 m m H g (placebo group).  Baseline pulse pressure: 57.8  m m H g (verap. 60 mg/day group), 52.8 m m H g (verap. 120 mg/day group), 56.3 mm H g (verap. 240 mg/day group), 52.2 mm H g (verap. 480 mg/day group), and 55.5 mm H g (placebo group). Interventions: Verapamil-SR 60 mg/day, 120 mg/day, 240 mg/day, 480 mg/day, or placebo. Primary a n d secondary outcomes: Change from baseline in supine peak and trough SBP, D B P (Instrument: not reported; by auscultation), FIR, E C G , lab tests. N o t e s : Trough BPs and associated SEs extracted from Table II. S D of B P change imputed from S D of endpoint BPs. Baseline H R s not reported. W D A E extracted from Table 111. Author deceased. F u n d i n g source: N o t reported. Messerli 1998 (210)  Design: M C / R / D B / P C  trial.  Placebo run-in period: 4 weeks.  Treatment duration: 6 weeks. Country: U S A Quality: Cochrane method = B; Jadad score = 3 Participants: D B P 95-114 m m H g for inclusion into trial. N = 631: 401 m, 230 f. Mean age: 54.5 years (range: 21-88).  127  Baseline  sitting  B P : 151.1/100.8  mm  153.6/100.5 m m H g (placebo group).  Hg  (verapamil  group),  Baseline pulse pressure: 50.3  m m H g (verapamil group), and 53.1 m m H g (placebo group). Interventions:  Verapamil SR 240 mg/day, trandolopril 4 mg/day,  verapamil SR 240 mg/trandolopril 4 mg combination, or placebo. Primary and secondary outcomes: Change from baseline in trough sitting D B P , SBP (Instrument: standard H g sphygmomanometer), H R at peak and trough; % responders ( D B P < 90 mm H g or > 10 mm H g decrease). N o t e s : Change in BPs and associated SEs extracted from Table 2. Changes in B P reported were already adjusted for placebo effect. Therefore, SDs of change in B P for the placebo group were imputed from other trials. F u n d i n g source: K n o l l Pharmaceutical Company, New Jersey. Neutel 1996 (211)  D e s i g n : M C / R / D B / P C trial. Washout period: 1-2 weeks. Placebo run-in period: 2-4 weeks. Treatment duration: 4 weeks. Country: U S A Quality: Cochrane method = B ; Jadad score = 3 Participants:  Sitting D B P 95-114 m m H g and mean  daytime  ambulatory D B P >90 mm H g for inclusion into trial. N = 95: 64m, 31 f. Mean age: 56.7 years (range: not reported). Baseline B P : 158.6/99.9 m m H g (verapamil group), 157.5/101.2 mm H g (placebo group). Baseline pulse pressure: 58.7 mm H g (verapamil group), and 56.3 m m H g (placebo group). Interventions: Verapamil C O E R - 2 4 240 mg/day, or placebo. Primary a n d secondary outcomes: Change from baseline in sitting SBP, D B P (Instrument: H g sphygmomanometer), A B P M (Instrument: SpaceLabs 90202). N o t e s : Change in BPs and associated SDs, W D A E extracted from text, p. 1203. Baseline H R s not reported. F u n d i n g source: N o t reported.  128  Neutel 1999 (212)  D e s i g n : M C / R / D B / P C trial. Washout period: duration not reported. Placebo run-in period: 2-4 weeks. Treatment duration: 8 weeks. Country: U S A Quality: Cochrane method = B; Jadad score = 3 Participants: Sitting D B P 95-114 m m H g and ambulatory D B P 90114 mm H g for inclusion into trial. N = 116: 75 m, 41 f. Mean age: 52.1 years (range: not reported). Baseline BP: 154.9/98.1 mm H g (verapamil group), 155.5/95.9 m m H g (placebo group).  Baseline pulse pressure: 56.8 m m H g (verapamil  group), and 59.6 mm H g (placebo group). Interventions: CODAS-verapamil 200 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in sitting SBP, D B P (Instrument: H g sphygmomanometer); A B P M (Instrument: Spacelabs 90207). N o t e s : Change in trough BPs and associated SDs extracted from text, p. 577. W D A E s extracted from text, p. 576. F u n d i n g source: N o t reported. Scholze 1998 (213)  D e s i g n : M C / R / D B / P C 3x4 factorial trial. Washout period: 1 week. Placebo run-in period: 4 weeks. Treatment duration: 6 weeks. Country: Germany Quality: Cochrane method — B; Jadad score = 3 Participants: Sitting D B P 100-115 m m H g for inclusion into trial. N = 424; ratio of m:f not reported.  Mean age and age range: not  reported. Baseline  B P : 167.1/105.9  mm  Hg  (ver.  120  mg/day  group),  169.9/106.9 mm Fig (ver. 180 mg/day group), and 171.4/105.4 m m H g (placebo group). Baseline pulse pressure: 61.2 mm Fig (ver. 120 mg/day group), 63 mm Fig (ver. 180 mg/day group), and 66 m m Fig (placebo group). Interventions: Verapamil SR 120 mg/day, 180 mg/day, trandolapril 0.5 mg/day, 1.0 mg/day, verapamil + trandolapril 120/0.5 mg/day,  129  120/1.0 mg/day, 180/0.5 mg/day, 180/1.0 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in trough sitting/standing SBP, D B P (fnstrument: H g sphygmomanometer), H R , routine haematological, biochemical and urinary investigations. N o t e s : Change in sitting BPs extracted from Table 1. Author contacted: baseline B P s / H R s , change in H R , and SDs of B P / H R change obtained. F u n d i n g source: K n o l l A . G . , Germany. Smith 2001 (214)  D e s i g n : M C / R / D B / P C trial.  Washout period: 1-14 days. Placebo  run-in period: 2-4 weeks. Treatment duration: 8 weeks. Country: U S A Quality: Cochrane method = B; Jadad score = 4 Participants: Sitting D B P >95 and <115 m m H g for inclusion into trial. N = 277: 177 m, 100 f. Mean age: 53.4 years (range: not reported). Baseline B P (ABPM): 148.0/93.3: mm H g (ver. 100 mg/day group), 150.1/93.4 (ver. 200 mg/day group), 148.6/92.6 (ver. 300 mg/day group), 146.9/92.4 (ver. 400 mg/day group), and 150.0/92.9 m m H g (placebo group).  Baseline pulse pressure: 54.7 m m Fig (ver. 100  mg/day group), 56.7 m m H g (ver. 200 mg/day group), 56 m m H g (ver. 300 mg/day group), 54.5 m m H g (ver. 400 mg/day group), and 57.1 m m H g (placebo group). Interventions:  CODAS-veraparnil 100 mg/day, 200 mg/day, 300  mg/day, 400 mg/day, or placebo. Primary a n d secondary outcomes: Change from baseline in trough sitting SBP, D B P (Instrument: not reported); A P B M  (fnstrument:  Spacelabs 90207). N o t e s : Change in office BPs and associated SDs extracted from text, page 17-18. W D A E extracted from text, p. 18. F u n d i n g source: N o t reported. Veratran 1997 (215)  Design: M C / R / D B / P C  trial.  Placebo run-in period: 4 weeks.  130  Treatment duration: 8 weeks. Country: Italy Quality: Cochrane method = B ; Jadad score = 3 Participants: Sitting D B P 100-110 m m H g for inclusion into trial N = 272 (234 in efficacy analysis): 134 m, 100 f. Mean age: 49.9 years (range: 18 — 65 for inclusion). Baseline BP: 156.0/104.2 mm H g (verapamil group), 158.2/103.5 m m H g (placebo group). Baseline pulse pressure: 51.8 m m H g (verapamil group), and 54.7 m m H g (placebo group). Interventions:  Verapamil SR 180 mg/day, trandolapril 1 mg/day,  verapamil SR 180 mg/day + trandolapril 1 mg/day combination, or placebo. Primary and secondary outcomes: Change from baseline in sitting SBP, D B P (Instrument: H g sphygmomanometer) and H R Notes: Week 8 BPs, H R s and associated SDs extracted from Table 1. SDs of B P / H R change imputed from week 8 B P s / H R s . # of patients in each group reported for efficacy analysis only — cannot use W D A E data since # of patients at randomization not reported. Funding source: N o t reported. V o n Manteuffel 1995 (216)  Design: M C / R / D B / P C  trial.  Placebo run-in period: 4 weeks.  Treatment duration: 6 weeks. Country: Germany and Austria Quality: Cochrane method = B; Jadad score = 2 Participants: Sitting D B P 100-115 m m H g for inclusion into trial N  = 176: 91 m, 85 f.  Mean age: 56.3 years (range: 1 8 - 7 0 for  inclusion). Baseline BP: 173.7/106.7 mm H g (verapamil group), 168.2/105.6 m m Fig (placebo group). Baseline pulse pressure: 67 m m H g (verpamail group), and 62.6 m m Tig (placebo group). Interventions: Verapamil SR 240 mg/day, hydrochlorothiazide 12.5 mg/day,  verapamil SR 240 mg/day  +  HCTZ  12.5 mg/day  131  combination, or placebo. Primary a n d secondary outcomes: Change from baseline in sitting SBP, D B P (Instrument: not reported), % responders (decrease in D B P > 10 m m H g or normalization of D B P to < 90 m m Hg). N o t e s : Published in German. Weighted mean SBPs at week 4 and 6 extracted from Figure 2. SD of change in SBP imputed from baseline SBP.  D B P at week 6 extracted from text, p. 375. S D of change in  D B P imputed from other trials. F u n d i n g source: N o t reported. White 1995 (217)  D e s i g n : M C / R / D B / P C trial. Washout period: 1 week. Placebo runin period: 2-4 weeks. Treatment duration: 8 weeks. Country: U S A Quality: Cochrane method = B; Jadad score = 2 Participants: Sitting D B P 95-114 m m Fig and A B P M daytime D B P > 90 mm H g for inclusion into trial. N = 287; ratio of m:f not reported. Mean age: 53.2 years (range: not reported). Baseline BP: 159/101 mm H g (ver. 120 mg/day group), 155/100 mm H g (ver. 180 mg/day), 158/102 mm Fig (ver. 360 mg/day), 158/101 mm H g (ver. 540 mg/day), 156/101 m m H g (placebo group). Baseline pulse pressure: 58 m m H g (ver. 120 mg/day group), 55 m m H g (ver. 180 mg/day group), 56 mm H g (ver. 360 mg/day group), 57 m m H g (ver. 540 mg/day group), and 55 m m H g (placebo group). Interventions:  Physiologic pattern  release  (PPR)-verapamil 120  mg/day, 180 mg/day, 360 mg/day, 540 mg/day, or placebo. Primary and secondary outcomes: Change from baseline in sitting SBP,  D B P (Instrument:  Hg  sphygmomanometer),  HR;  ABPM  (Instrument: Spacelabs 90202). N o t e s : Change in trough clinic B P s / H R s and SEs extracted from Table II. W D A E data incompletely reported. F u n d i n g source: N o t reported.  132  \  White 2002  D e s i g n : M C / R / D B / P C trial. Washout period: 1-2 weeks. Placebo  (218)  run-in period: 2-4 weeks. Treatment duration: 8 weeks (forced titration at week 4). Country: U S A Quality: Cochrane method = B; Jadad score — 3 Participants: Sitting D B P 95-114 m m H g and A B P M daytime D B P > 85 m m Pig for inclusion into trial. N = 357: 219 m, 138 f. Mean age: 54.5 years (range: not reported). Baseline sitting BP: 146/95 mm H g (verapamil group), 148/95 m m H g (placebo group).  Baseline pulse pressure: 51 m m Pig (verapamil  group), and 53 m m H g (placebo group). Interventions: COER-verapamil 240  360 mg/day, enalapril l O ^ -  20 mg/day, losartan 50~^ 100 mg/day, or placebo. Primary  and secondary  outcomes:  Change  from  baseline in  sitting/standing SBP, D B P (fnstrument: H g sphygmomanometer) & H R ; A B P M (fnstrument: Spacelabs 90207). N o t e s : Change in sitting B P and associated SDs at week 8 extracted from Table IV. Clinic H R s not reported.  W D A E ' reported but on  what dose these occurred is not specified. F u n d i n g source: N o t reported.  133  3.3 Characteristics of studies awaiting data from authors T a b l e 31: Characteristics of studies meeting inclusion criteria a n d awaiting office b l o o d pressure data from authors Study Bakris 2002 (219)  Study Description D e s i g n : M C / R / D B / P C trial. Washout period: 1-2 weeks. Placebo run-in period: 2-4 weeks. Treatment duration: 8 weeks (forced titration at week 4) Country: U S A and Canada Quality: Cochrane method = B; Jadad score = 3 Participants: Sitting D B P 95-114 mm H g for inclusion into trial. N = 405: 223 m, 182 f. Mean age: years (inclusion range: 21 — 80). Baseline sitting B P : 145/95 mm H g (verap. group), and 146/96 m m H g (placebo group). Baseline pulse pressure: 50 m m H g (verap. group), and 50 m m H g (placebo group). Interventions:  COER-verapamil 240  20 mg/day, losartan 50 Primary  360 mg/day, enalapril 10-^  100 mg/day, or placebo.  a n d secondary  outcomes:  Change  from  baseline in  sitting/standing office SBP, D B P (fnstrument: not reported), 24-hour A B P M (fnstrument: Spacelabs 90207), rate of rise of B P / H R during the morning acceleration phase. N o t e s : Office B P s / H R s to be obtained from author. F u n d i n g Source: Pharmacia Corp., U S A . Glasser 2003 (220)  Design: M C / R / D B / P C  trial.  Placebo run-in period: 3-4 weeks.  Treatment duration: 7 weeks. Country: U S A Quality: Cochrane method = B; Jadad score = 3 Participants: sitting D B P 100-114 mm H g for inclusion into trial. N — 478: 303 m, 175 f. Mean age: 52 years (inclusion range: 18 — 70). Baseline BP/pulse pressure: office values not reported. Interventions: Diltizaem extended-release 120 mg/day, 240 mg/day, 360 mg/day, 540 mg/day, or placebo.  134  Primary  a n d secondary  outcomes:  Change  from  baseline in  sitting/standing office SBP, D B P (Instrument: not reported) and H R ; 24-hour A B P M (Instrument: Spacelabs 90207). N o t e s : Office B P s / H R s to be obtained from author. F u n d i n g Source: N o t reported.  135  3.4 Characteristics of excluded studies Table 32 below documents the reasons why 134 studies that met the primary inclusion criteria had to be excluded from the analysis. Several of these studies were either cross-over trials that did not report pre-crossover data or parallel-group studies that allowed dose titration in non-responders.  Other common reasons for exclusion included failure to  report details such as number of patients in each arm, baseline blood pressures, or pretitration data. T a b l e 32: Reasons for exclusion of certain studies meeting inclusion criteria Study I D Abadie 1985 (221) Ahmed 1992a (222) Allrkmets 1997 (223) Andersson 1989 (224) [Duplicate publication: Andersson 1990 (225)1 Andren 1988 (226) Arita 1999 (227) ArziUi 1993 (228) Bainbridge 1993 (229) B aylac-D omengetroy 1990 (230) BeUet 1985 (231) Bossini 1990 (232)  Reason for exclusion R / D B / P C parallel group trial that assessed the effect of nifedipine 30 mg/day on carbohydrate metabolism — no blood pressures reported. R / D B / P C cross-over trial with no pre-crossover data reported for the 1" four weeks of treatment (nicardipine 90 mg/day vs. placebo) R / D B / P C parallel group trial with no data reported for the 1 four weeks of treatment prior to titration in non-responders (isradipine 5 mg/day vs. placebo) R / D B / P C cross-over trial with no pre-crossover data reported for the 1 nine weeks of treatment [isradipine 2.5 mg (3 weeks) 5 mg (3 weeks) 7.5 mg (3 weeks) vs: placebo] st  st  R / D B / P C cross-over trial with no data reported for the 1 three weeks of treatment, prior to cross-over and also prior to titration in non-responders (diltiazem 240 mg/day vs. placebo) R / D B / P C cross-over trial with no pre-crossover data for the 1 four weeks of treatment (azelnidipine 8 mg vs. placebo) R / D B / P C cross-over trial with no pre-crossover data for the 1 4 weeks of treatment (isradipine S R O 5 mg vs. placebo) R / D B / P C cross-over trial with no pre-crossover data for the 1 4 weeks of treatment (felodipine E R 5 mg/day vs. ramipril 2.5 mg/day vs. felodipine E R 5 mg + ramipril 2.5 mg vs. placebo) R / D B / P C parallel group trial that studied nitrendipine 20 mg/day vs. placebo. N o data reported for the placebo group. st  st  st  st  R / D B / P C parallel group trial. Dose titration in non-responders after 15 days of treatment (nicardipine 60 mg/day vs. placebo) R / D B / P C cross-over trial with no pre-crossover data for the 1 4 weeks of treatment (felodipine E R 10 mg/day vs. placebo) st  136  Study I D British Isradipine Hypertension Group 1989 (233) CampbeU 1990 (234) Capewell 1989 (235) Carr 1990 (236)  Chalmers 1990 (237) Clement 1987 (238) Cleroux 1992 (239) Cleroux 1994 (240)  Cox 1988 (241) Cox 1989 (242) Crozier 1990 (243) De Brmjn 1988 (244) Diemont 1991 (245) Dittrich 1992 (246) Draaijer 1995 (247)  Reason for exclusion R / D B / P C parallel group trial with no data reported for the 1 three weeks of treatment prior to titration in non-responders [isradipine 2.5 mg/day (1.25 mg twice daily or 2.5 mg once daily) vs. placebo]. R / D B / P C parallel group trial with no data reported for the 1 six weeks of treatment prior to titration in non-responders (felodipine E R 5 mg vs. placebo) R / D B / P C cross-over trial with no pre-crossover data for the 1 4 weeks of treatment (felodipine 10 mg/day vs. placebo). Titration in non-responders after 2 weeks treatment. R / D B / P C parallel group trial with initial dose titrated to response. Data from responders and non-responders reported separately, (isradipine 5, 10, 15, or 20 mg, depending on B P response vs. placebo). Reports on 3 studies. The 3 study was a R / D B / P C cross-over trial with no pre-crossover data for the 1 four weeks of treatment with diltiazem 240 mg/day R / D B / P C cross-over trial with no pre-crossover data for the 1 three weeks of treatment (nifedipine SR 40 mg/day vs. atenolol 100 mg/day vs. placebo). N o baseline B P values given. R / D B / P C cross-over trial with no pre-crossover data for the 1 four weeks of treatment (isradipine SR 5 mg vs. placebo) R / D B / P C cross-over trial with no pre-crossover data for the 1 four weeks of treatment. 8-week treatment periods but titration in non-responders after 4 weeks treatment, (verapamil 240mg/day vs. quniapril 10 mg/day vs. atenolol 50 mg/day) R / D B / P C cross-over trial with no pre-crossover data for the 1 six weeks of treatment (verapamil 40/80/120 mg/day vs. placebo). Open titration phase to determine dose before randomization. R / D B / P C cross-over trial with no pre-crossover data for the 1 four weeks of treatment (nicardipine 120 mg/day vs. placebo) R / D B / P C cross-over trial with no pre-crossover data for the 1" five weeks of treatment (felodipine 20 mg/day vs. placebo) R / D B / P C parallel group trial with 8-week treatment period. Dose titration in non-responders at 2-week intervals (starting dose: amlodipine 2.5 mg vs. atenolol 50 mg/day vs. placebo) Reports on two studies: one single blind study and one cross-over trial with no pre-crossover data for the 1 four weeks of isradipine M R 5 mg/day vs. isradipine M R 10 mg/day vs. placebo R / D B / P C parallel group trial. Doses of 60 mg, 90 mg, or 120 mg/day of nicardipine were studied. Number of patients on each dose and titration schedule not reported. R / D B / P C parallel group trial. B P data reported only at the end of six months of treatment (felodipine 5 mg/day vs. enalapril 10 mg/day vs. placebo) st  st  st  rd  st  st  st  st  st  st  st  Study I D Dupont 1991 (248) . Duprez 1991 (249) Durel 1992 (250)  Reason for exclusion R / D B / P C cross-over trial with no pre-crossover data reported for the 1 three weeks of diltiazem L P 300 mg/day vs. placebo R / D B / P C cross-over trial with no pre-crossover data reported for the 1 six weeks of isradipine 10 mg/day vs. placebo A study consisting of six 28-day cross-over trials in a quasi-random order. N o baseline B P data and no pre-crossover data reported for the 1 4 weeks of verapamil 240 mg/day vs. placebo. R / D B / P C parallel group trial with titration in non-responders after 15 days of treatment with nicardipine 60 mg/day vs. placebo. st  st  st  Faguer de Moustier 1989 (251) [Duplicate publication: Faguer de Moustier 1990 (252) Ferreira-Filho 1995 (253) Fogan 1993 (254) Forette 1984 (255) [Duplicate publication: Forette 1985 (256)] Fnshman 1988 (257) (Multiple publications: Frishman 1994 (258) , Johnson 1992 (259), Johnson 1995 (260)] Gavras 1987 (261) Gebera 1996 (262) G r i m m 2002 (263) Hamilton B 1987 (264) Harrington 1987 (265)  R / D B / P C cross-over trial with no pre-crossover data reported for the 1 four weeks of nifedipine SR 40 mg/day vs. placebo R / D B / P C cross-over trial with no pre-crossover data reported for the 1 four weeks of isradipine S R O 5 mg/day (administered morning or night) vs. placebo R / D B / P C parallel group trial with titration in non-responders every 8 days during treatment, starting with nicardipine 30 mg/day vs. placebo. st  st  R / D B / P C parallel group trial with titration in non-responders every 2 weeks during treatment, starting with amlodipine 2.5 mg/day vs. atenolol 50 mg vs. placebo.  R / D B / P C parallel group trial with titration in non-responders after 1 week of treatment with nifedipine GITS 30 mg/day vs. placebo. R / D B / P C cross-over trial with dose titration in non-responders after 2 weeks of treatment with verapamil 240 mg/day vs. placebo. Pre-crossover data not reported R / D B / P C parallel group trial with no data reported for the 1 four weeks of treatment with amlodipine 5 mg/day vs. placebo prior to titration in non-responders. Reports on several phase III studies, including a R / D B / P C parallel group trial of isradipine 5 mg/day, 10 mg/day, 15 mg/day, 20 mg/day, or placebo. B P data not reported. Reports on 2 studies: a short-term single-blind study, and a R / D B / P C cross-over trial with dose titration in non-responders every 2 weeks of treatment, starting with felodipine 5 mg/day vs. placebo. Pre-crossover data not reported. st  138  Study ID Hedback 1984 (266) Herrera 1997 (267) Honorato 1989 (268) Hosie 1991 (269) [Conference abstract: Hosie 1990 (270)] Jeffrey 1990 (271)  Kelemen 1990 (272) [Duplicate publication: Stewart 1990 (273)] Khalil-Manesh 1987 (274) Kjellstrom 1994 (275) Klauser 1990 (276) [Duplicate publication: Klauser 1991 (277)] Krakoff 1989 (278) Lacourciere 1995 (279) Lacourciere 1998 (280) Landmark 1985 (281) Lessem 1990 (282)  Reason for exclusion R / D B / P C cross-over trial with no pre-crossover data reported for the 1 four weeks of verapamil 360 mg/day vs. placebo R / D B / P C parallel group trial with titration in non-responders during weeks 2-4 of treatment, starting with diltiazem C D 180 mg/day vs. placebo. R / D B / P C parallel group 6-week trial of nitrendipine 20 mg/day vs. placebo. N o " n " values reported. Author contacted — no response. R / D B / P C parallel group trial with titration in non-responders after 2 week of treatment with felodipine E R 5 mg/day vs. placebo. st  R / D B / P C cross-over trial with dose titration in non-responders after 2 weeks of treatment with felodipine 10 mg/day vs. placebo. N o pre-crossover data reported. Only mean arterial pressure reported. R / D B / P C parallel group trial with exercise co-intervention.  R / D B / P C parallel group trial with titration in non-responders every 2 weeks during treatment, starting with diltiazem 120 mg/day vs. placebo. R / D B / P C cross-over trial with dose titration in non-responders after 2 weeks of treatment with felodipine E R 10 mg/day vs. placebo. Baseline B P and pre-crossover data not reported R / D B / P C cross-over trial with 4-week treatment periods. Forced titration with isradipine 10 mg/day (2 weeks) isradipine 20 mg/day (2 weeks) vs. placebo. N o pre-crossover data reported  R / D B / P C parallel group trial with titration in non-responders after 2 weeks of treatment with nicardipine 60 mg/day vs. placebo. R / D B / P C parallel group trial with an 8-week treatment period during which there was forced titration every 2 weeks, starting with diltiazem E R 120 mg/day. R / D B / P C parallel group trial with no data reported for the 1 eight weeks of treatment with amlodipine 5 mg/day vs. placebo prior to titration in non-responders. R / D B / P C cross-over trial with no pre-crossover data reported for the 1 eight weeks of nifedipine 40 mg/day vs. placebo R / D B / P C parallel group trial that studied nicardipine 30 mg/day, 60 mg/day, 90 mg/day, and 120 mg/day vs. placebo. N o baseline BPs reported. st  st  139  Study I D Letzel 1990 (283) Lewis 1978 (284) Lok 1989 (285) LOMIR-MCT-IL 1993 (286) [Duplicate publication: Yodfat 1996 (287)1 Lorimer 1988 (288)  Reason for exclusion R / D B / P C parallel group trial with no baseline B P data reported (verapamil 160 mg/day vs. H C T Z 25 mg/day, vs. triamterene 50 mg/day vs. all possible combinations vs. placebo). R / D B / P C cross-over trial with no pre-crossover data reported for the 1 four weeks of verapamil 240 mg/day vs. verapamil 360 mg/day vs. placebo R / D B / P C parallel group trial with titration in non-responders after 2 weeks of treatment with felodipine 5 mg/day vs. placebo R / D B / P C parallel group trial that does not report number of patients at the 1 four weeks of treatment with isradipine 2.5 mg/day, prior to titration in non-responders. Author contacted; data no longer available. st  st  R / D B / P C parallel group trial with titration in non-responders every 2 week of treatment, starting with amlodipine 2.5 mg/day vs. verapamil 160 mg/day vs. placebo.  [Duplicate publication: Lorimer 1989 (289)1 L o w 1993 R / D B / P C cross-over trial with no pre-crossover data reported for (290) the 1 four weeks of isradipine 5 mg/day vs. placebo. Lyons 1994 R / D B / P C parallel group trial with titration in non-responders after (291) 2 weeks of treatment with amlodipine 5 mg/day vs. placebo. Macphee 1989 R / D B / P C cross-over trial with no pre-crossover data reported for (292) the 1 four weeks of nifedipine 20 mg/day vs. placebo. Mancia 1992 R / D B / P C parallel group trial of verapamil SR 240 mg/day, (293) nitrendipine 20 mg/day, enalapril 20 mg/day vs. placebo. Only mean 24-hour A B P M measurements reported. Author contacted — no reponse. Massie 1992 R / D B / P C parallel group trial with titration in non-responders (294) every 2 week of treatment, starting with diltiazem C D 120 mg/day vs. placebo. Mattarei 1987 R / D B / P C cross-over trial with titration in non-responders (295) (titration schedule not reported), starting with nifedipine 20 mg/day vs. placebo. N o pre-crossover data reported. Mclnnes 1985 R / D B / P C cross-over trial with no pre-crossover data for the 1 four weeks of treatment (verapamil 360 mg/day vs. propranolol (296) 240 mg/day vs. verapamil 360 mg/day + propranolol 240 mg/day [Duplicate publication: vs. placebo). Dargie 1986 (297)1 Megnien 1995 R / D B / P C parallel group trial with titration in non-responders after (298) 2 weeks of treatment with amlodipine 5 mg/day vs. placebo st  st  st  140  Study I D Midtbo 1980 (299) [Duplicate publication: Storstein 1981 (300)1 Morgan 2001 (301)  Morgan 2002 (302) Moser 1984 (303) Nelson 1986 (304) Neutel 1995 (305) Nikkila 1989 (306) Nikolova 1995 (307)  Nilsson 1996 (308)  Ollivier 1995 (309) Pacheco 1986 (310) Pandita-Gunawardena 1999 (311) Pannarale 1996 (312)  Reason for exclusion R / D B / P C cross-over trial with no pre-crossover data reported for the 1 four weeks of verapamil 480 mg/day vs. placebo st  R / D B / P C cross-over trial with forced titration after 1 month of treatment. N o pre-crossover data for the 1 2 months of treatment with felodipine (dose not reported) vs. amlodipine 5 mg (1 month) amlodipine 10 mg (1 month) vs. placebo. B P data from felodipine and amlodipine treatments lumped together. R / D B / P C cross-over trial with no pre-crossover data reported for the 1" four weeks of felodipine 5 mg/day vs. placebo. R / D B / P C parallel group trial with a weekly forced titration for 3 weeks, starting with nitrendipine 10 mg/day vs. placebo, followed by a 2-week maintenance period. R / D B / P C parallel group trial with 3-week treatment period and weekly forced titration: isradipine 5 mg/day (1 week) - ^ 1 0 mg/day (1 week) 20 mg/day (1 week) vs. placebo. R / D B / P C parallel group trial with no data reported for the 1 four weeks of treatment with verapamil 120 mg/day vs. placebo prior to titration in non-responders. Author contacted: no response? R / D B / P C cross-over trial with no pre-crossover data reported for the 1 three weeks of diltiazem 240 mg/day vs. placebo, prior to titration in non-reponsders. R / D B / P C parallel group trial comparing nifedipine 30 mg/day with placebo in treatment of hypertensive venous leg ulcers. Patients in both groups were allowed centrally acting antihypertensive drugs in addition to randomized treatment. R / D B / P C cross-over trial with single-blind diltiazem O D 300 mg/day run-in prior to randomization. N o pre-crossover data reported for the 1" six weeks of diltiazem O D 300 mg/day vs. placebo. R / D B / P C parallel group trial with titration in non-responders every 2 weeks, starting with diltiazem SR 200 mg/day vs. placebo. R / D B / P C parallel group trial with titration in non-responders during the 1 three weeks of treatment, starting with nifedipine 30 mg/day vs. placebo D B / P C parallel group trial. Allocation by minimization. N o data reported for the 1 four weeks of treatment with amlodipine 5 mg vs. placebo) prior to titration in non-responders R / D B / P C cross-over trial with no pre-crossover data reported for the 1 four weeks of felodipine E R 10 mg/day vs. placebo st  st  st  st  st  st  141  Study ID Persson 1989a (313) [Multiple publications: Persson 1989b (314), Persson 1992 (315)] Pool 1985 (316) Pool 1986 (317) Pool 1988 (318) [Duplicate publication: Pool 1989 (319)1 Pool 1990 (320) Pnsant 1998 (321) [Duplicate publication: Prisant 1999 (322)] Ramirez 1992 (323)  Ricciardelli 1997 (324) Romero-Vecchione 1995 (325) Rossi 2002 (326) Salvetti 1987 (327)  Salvetti 1989a (328)  Reason for exclusion R / D B / P C cross-over trial with no data reported for the 1 three weeks of treatment with isradipine 5 mg/day vs. placebo prior to titration in non-responders. N o pre-crossover data reported. st  R / D B / P C parallel group trial with titration in non-responders every 2 weeks of treatment, starting with diltiazem 120 mg/day vs. placebo. R / D B / P C parallel group trial with titration in non-responders every 2 weeks of treatment, starting with diltiazem 120 mg/day vs. placebo. R / D B / P C parallel group trial with titration in non-responders every 2 weeks of treatment, starting with diltiazem 120 mg/day vs. placebo.  Reports on 3 trials. Trial #2 was a R / D B / P C parallel group trial with no data reported in the 1 4 weeks of treatment with diltiazem 240 mg/day vs. placebo prior to titration in non-responders . R / D B / P C parallel group trial with a 6-week titration period during which doses were titrated in non-responders every 2 weeks, starting with amlodipine 2.5 mg/day vs. placebo. st  R / D B / P C parallel group trial that studied nifedipine GfTS 60 mg/day vs. 90 mg/day vs. 120 mg/day vs. 180 mg/day vs. placebo. BPs from nifedipine arms were grouped together and not reported for each dose separately. R / D B / P C parallel group trial with no data reported for the 1 six weeks of treatment with nitrendipine 10 mg/day vs. placebo prior to titration in non-responders. D B / P C parallel group trial with titration in non-responders every week for 4 weeks, starting with nifedipine 10 mg/day. st  R / D B / P C cross-over trial with no pre-crossover data reported for the 1 three weeks of lacidipine 4 mg/day vs. placebo. R / D B / P C cross-over trial with no baseline data and no precrossover data reported for the 1 four weeks of nifedipine 40 mg/day vs. placebo. Only mean arterial blood pressure values given. R / D B / P C cross-over trial with no pre-crossover data reported for the 1 four weeks of nicardipine SR 80 mg/day vs. placebo. st  st  st  142  Study I D Salvetti 1989b (329) [Duplicate publication: Salvetti 1991 (330)1 Salvetti 1996 (331) Schuster 1998 (332) Siche 2001 (333) Simon 1984 (334) Slomm 1991 (335) Spence 2000 (336) Spieker 1998 (337) [Duplicate publication: Smilde 2000 (338)] Spritzer 1990 (339) Staessen 1989 (340) Stornello 1989 (341) Stornello 1990 (342) Suzuki 1999 (343) Szlachic 1989 (344) Tonkin 1990 (345)  Reason for exclusion R / D B / P C cross-over trial with no baseline S B P / D B P data and no pre-crossover data reported for the 1 four weeks of nifedipine 40 mg/day vs. placebo. st  R / D B / P C cross-over trial with no pre-crossover data reported for the P ' four weeks of nifedipine GfTS 30 mg/day vs. nifedipine retard 40 mg/day vs. placebo. R / D B / P C cross-over trial with dtradon in non-responders after 2 weeks of treatment with felodipine 5 mg/day vs. placebo. N o precrossover data reported. R / D B / P C parallel group trial with no data reported for the f four weeks of treatment with amlodipine 5 mg/day vs. placebo prior to titration in non-responders. R / D B / P C parallel group trial with titration in non-responders every week for 3 weeks, starting with nitrendipine fO mg/day vs. placebo. R / D B / P C parallel group trial with titration in non-responders, starting with isradipine 2.5 mg/day vs. placebo. Timing of titration and number of patients on each dose are not reported. R / D B / P C cross-over trial with no pre-crossover data reported for the 1 four weeks of amlodipine 5 mg/day titrated up to 10 mg/day vs. placebo. Presents results of 4 studies, one of which was a R / D B / P C crossover trial with no pre-crossover data reported for barnidipine 20 mg vs. placebo. st  st  R / D B / P C cross-over trial with no pre-crossover data reported for the 1 four weeks of nitrendipine 20 mg/day vs. placebo. R / D B / P C cross-over trial with no pre-crossover data reported for the 1 four weeks of isradipine 10 mg/day vs. placebo, prior to titration in non-responders. R / D B / P C cross-over trial with no pre-crossover data reported for the 1 four weeks of nicardipine 60 mg/day vs. placebo. R / D B / P C cross-over trial with no pre-crossover data reported for the 1 six weeks of nicardipine-SR 80 mg/day vs. placebo. R / D B / P C parallel group trial with titration in non-responders, starting with benidipine 4 mg/day vs. placebo. Timing of titration and number of patients on each dose are not reported. R / D B / P C parallel group trial with an 8-week titration in nonresponders, starting with diltiazem 120 mg/day vs. placebo R / D B / P C cross-over trial with no pre-crossover data reported for the 1 four weeks of diltiazem 240 mg/day vs. placebo st  st  st  st  st  143  Study I D Tschoepe 1992 (346) Van Bortel 1990 (347) Van Merode 1989 (348) [Duplicate publication: Van Merode 1990 (349)1 Vanhees1991 (350) Viskoper 1991a (351) [Duplicate publications: Viskoper 1991b (352), Viskoper 1992 (353)1 Watts 1998 (354) Weber 1988 (355) Webster 1987 (356) [Duplicate publication: Webster 1988 (357)] Webster 1989 (358) Webster 1991 (359) Weir 1998 (360) [Duplicate publiction: Chrysant 2000 (361)] Welzel 1990 (362)  Reason for exclusion R / D B / P C parallel group trial that studied nitrendipine 20mg/day vs. placebo. BPs reported as % change from pretreatment values, but no baseline values are given. R / D B / P C cross-over trial with no pre-crossover data reported for the 1 four weeks of verapamil 360 mg/day vs. placebo R / D B / P C cross-over trial with no pre-crossover data reported for the 1 four weeks of verapamil 360 mg/day vs. placebo. N o baseline BPs reported. st  st  R / D B / P C cross-over trial with no pre-crossover data reported for the 1 three weeks of verapamil 240 mg/day vs. placebo. R / D B / P C parallel group trial with no data reported for the 1 four weeks of treatment with isradipine 5 mg/day vs. placebo prior to titration in non-responders. st  st  R / D B / P C cross-over trial with titration in non-responders after 2 weeks of treatment with amlodipine 5 mg/day vs. diltiazem C D 180mg/day vs. placebo. N o pre-crossover data reported. R / D B / P C parallel group trial with titration in non-responders, starting with diltiazem 240 mg/day vs. placebo. Timing of titration and number of patients on each dose are not reported. R / D B / P C parallel group trial with forced titration amlodipine 2.5 mg/day (2 weeks) 5 mg/day (2 weeks) - ^ 1 0 mg/day (4 weeks) vs. placebo. Dosage could be reduced if there was excessive lowering of B P or adverse effects from therapy. Data reported excludes patients whose time since the last dose at the 8-week visit was outside of the 18-30 hour "window". R / D B / P C cross-over trial with no pre-crossover data reported for the 1 four weeks of nicardipine 120 mg/day vs. placebo. R / D B / P C cross-over trial with no pre-crossover data reported for the 1 four weeks of nicardipine (standard formulation) 90 mg/day vs. nicardipine-SR 120 mgvs. placebo. R / D B / P C parallel group trial with dietary salt co-intervention with low (5-10 mg/day) and high doses (10-20 mg/day) of isradipine. st  st  R / D B / P C parallel group trial with no data reported for the 1 three weeks of treatment with isradipine 2.5 mg/day vs. nifedipine retard 40 mg/day vs. placebo prior to titration in non-responders. st  Study I D White 1997 (363) Wilson 1989 (364) Winer 1987 (365) Winer 1990 (366) Wing 1994a (367)  Wing 1994b (368)  Wing 1997 (369) Woehler 1992 (370)  Wysocki 1992 (371) [Duplicate publication: Wysocki 1998 (372)] Yamakado 1993 (373) Yasky 1991 (374) Zito 1991 (375)  Reason for exclusion R / D B / P C parallel trial with 8 week treatment period of verapamil C O E R 120, 180, 360, 540 mg/day vs. placebo. A B P M study of dippers vs. non-dippers (no office pressures reported). R / D B / P C cross-over trial with no pre-crossover data reported for the 1 six weeks of nisoldipine 20 mg/day vs. placebo. BPs reported as mean arterial BPs. R / D B / P C parallel group trial with 4-week treatment period during which non-responders were titrated in non-responders every week, starting with isradipine 5 mg/day vs. placebo. R / D B / P C parallel group trial with 6-week treatment period during which doses were titrated in non-responders every week up to week 4, starting with isradipine 5 mg/day vs. placebo. R / D B / P C cross-over trial with 6-week treatment periods, comprised of a forced titration after 1 week of treatment with felodipine 5 mg/day vs. placebo, followed by a titration in nonresponders at week 3. Doses could be stepped backwards i f unacceptable adverse drug effects were experienced. R / D B / P C cross-over trial with 6-week treatment periods and titration in non-responders after 2 and 3 weeks of treatment, starting with feldopine 5 mg/day. Doses could be stepped backwards i f unacceptable adverse drug effects were experienced. R / D B / P C cross-over trial with titration in non-responders after 2 weeks of treatment with lacidipine 2 mg/day vs. placebo. N o precrossover data reported. R / D B / P C parallel group trial with a 6-week treatment period during which there was forced titration every 2 weeks, starting with diltiazem 180 mg/day. Doses were not increased in patients whose D B P fell below 75 m m H g or those who had a significant adverse event. R / D B / P C cross-over trial with no data reported for the 1 three weeks of treatment with isradipine 5 mg/day vs. placebo prior to titration in non-responders. N o pre-crossover data reported. st  st  R / P C cross-over trial. Unclear i f double-blind. N o data reported for the 1 four weeks of treatment with diltiazem 180 mg/day vs. placebo. R / D B / P C cross-over trial with no pre-crossover data reported for the 1 four weeks of diltiazem 240 mg/day vs. placebo. R / D B / P C parallel group trial that studied lacidipine 2 mg/day vs. 4 mg/day vs. placebo but does not report results separately for each dosage group. st  st  145  3.5 Overview of included trials Summary data of baseline characteristics of the 106 included trials are listed in Table 33. The number of patients in the placebo group tends to be substantially smaller than the number in the C C B treatment groups because there are multiarm trials comparing several doses of active drug with one placebo arm.  T a b l e 3 3 : Overview of i n c l u d e d trials w i t h C C B s as monotherapy for primary hypertension Drug Dose range Total # of trials  Total # of patients in CCB treatment group  Total # of patients in placebo group  Mean age of patients (years)  Mean duration of treatment (weeks)  Amlodipine 2.5-10 mg/day 9 trials Barnidipine 10-30 mg/day 1 trial Darodipine 100-300 mg/day 1 trial Diltiazem 90-540 mg/day 12 trials  889  534  52.7  6.3  140  50  55.8  6  164.0/102.5; 61.5  33  10  51.7  4  156.9/103.0; 53.9  1545  641  55.4  5.8  154.3/100.7; 53.6  1190  556  55.5  6.3  158.2/100.1; 58.1  1091  404  55.0  4.9  161.0/105.2; 55.8  86  45  70.3  4  177.4/101.8; 75.6  Felodipine 2.5-20 mg/day 11 trials Isradipine 2.5-20 mg/day 11 trials Lacidipine 2-4 mg/day 1 trial  Mean baseline SBP/DBP; Pulse Pressure (mm Hg) 157.5/104.0; 53.5  146  Drug Dose range Total # of trials  Total # of patients in CCB treatment group  Total # of patients in placebo group  Mean age of patients (years)  Mean duration of treatment (weeks)  Mean baseline SBP/DBP; Pulse Pressure (mm Hg) 161.9/98.9; 63.1  Lercanidipine 2.5-20 mg/day 6 trials Lidoflazine 180 mg/day 1 trial Manidipine 10-40 mg/day 2 trials Mibefradil 6.25-200 mg/day 3 trials Nicardipine 40-100 mg/day 8 trials Nifedipine 20-100 mg/day 12 trials Nilvadipine 8-30 mg/day 2 trials Nisoldipine 10-30 mg/day 1 trial Nitrendipine 10-20 mg/day 10 trials Prandipine 1-8 mg/day 1 trial Tiapamil 300-1200 mg/day 1 trial Verapamil 60-540 mg/day 13 trials  532  252  57.7  3.8  5  5  43.8  12  138.0/95.0; 43.0  66  40  67.8  6  167.4/99.1; 68.3  685  128  60.0  4  175.7/101.3; 74.4  303  169  52.3  6.4  157.4/101.6; 55.9  675  413  52.7  7.1  157.1/102.1; 55.0  172  80  54.2  4  162.6/102.3; 60.3  148  58  52.1  6  161.4/106.9; 54.5  310  255  55.7  6.2  162.5/100.4; 62.1  144  32  52.0  4  154.6/102.8; 51.8  49  9  50.5  6  152.0/101.9; 50.2  1450  684  53.4  6  154.6/99.8; 54.8  TOTAL: 106 trials, 13 878 patients  9513  4365  55.0  5.7  158.2/101.6; 56.7  147  3.6 Values used to impute m i s s i n g variances 3.6.1.  Standard deviation of b l o o d pressure change  Thkty-seven (35%) of the trials included in this systematic review reported the standard deviation of the change in SBP and/or D B P .  These values were used to calculate  weighted means of the standard deviation of changes in SBP and D B P for both the C C B treatment and placebo groups. For the CCB-treatment group, the weighted mean standard deviations of the change in SBP and D B P were 13.3 m m H g (SD 2.1) and 7.7 m m Fig (SD 1.6), respectively. For the placebo group, the weighted mean S D of B P change was 13.0 m m H g (SD 4.1) for SBP and 7.4 mm H g (SD 2.3) for D B P . Those trials whose reported S D values were not witliLn three standard deviations of the weighted mean S D of B P change were then excluded from the calcuation in order to obtain an adjusted weighted mean S D of B P change.  Three trials with outlier SDs were  excluded from the calculation (Chan 1997, Farsang 2001, Asmar 1992), resulting in an adjusted weighted mean S D for C C B treatment of 13.5 m m H g and 7.8 m m H g for SBP and D B P , respectively, and corresponding values for placebo of 14.1 m m H g and 7.9 mm H g . There was no statistically significant difference in S D of SBP change between placebo and C C B treatment groups (p=0.15), nor in S D of D B P change (p=0.86). The adjusted values were used to impute the S D of B P change in the trials with outlier SDs and for other trials as per the protocol's imputation hierarchy. The S D of B P change was imputed for 70 (67%) of the included trials with extracted B P data. O f these, there were 3 trials (4%) in which variances were imputed by calculating the pooled standard deviation from a reported p value; 1 (1%) was imputed using S D reported from B P measured in a different position; 38 (56%) were imputed using endpoint  148  SDs; 6 (9%) were imputed using baseline SDs (for SBP only); 20 (30%) were imputed using the mean S D data reported in other trials.  3.6. 2. Standard deviation of heart rate change Twelve trials reported S D of the change in heart rate, resulting in a weighted mean value of 8.8 beats/minute for C C B group and 8.4 beats/minute for placebo group.  There  was no statistically significant difference between the S D of heart rate change in C C B - and placebo-treated groups (p=0.4). These values were used to impute the standard deviations according to the hierarchy outlined in the protocol. 3.7  Dose-related b l o o d pressure lowering of i n d i v i d u a l C C B drugs Effect sizes are expressed as the weighted mean difference (CCB treatment effect  minus placebo effect) followed by the 95% confidence interval in parentheses. analyses, "statistical significance" refers to p < 0.05.  In all  For analyses for which there was  statistically significant interstudy heterogeneity, the random effects model was applied to determine i f there was a statistically significant difference between C C B treatment and placebo. If a statistically significant difference was present using the random effects model, the effect size is reported using the fixed effect model because of the tendency of the former to overweight smaller trials. The weighted mean changes in SBP and D B P in placebo groups across all trials was -3.5 mm H g (range -16.7 - 9) and -3.8 m m Fig (range -11.4 - 4.5), respectively.  149  3.7.1  A m l o d i p i n e vs. placebo  T a b l e 34: B l o o d pressure lowering efficacy of amlodipine 1.25 - 10 m g / d a y . Fixed effect model with 95% confidence interval. Change in SBP, Change in D B P , Dose of # of trials Total # of mm H g patients in mm H g Amlodipine (95% CI) treatment group (95% CI) -1.44 2 88 -1.44 1.25 mg/day (-3.65, 0.78) (-5.46, 2.57) -5.30* -3.78* 2.5 mg/day 3 158 (-5.51, -2.05) (-8.58, -2.03) -10.72* -5.53* 556 5 mg/day 7 (-12.38, -9.05) (-6.50, -4.55) -7.29* 2 55 -10.71* 10 mg/day (-9.95,-4.62) (-15.70, -5.72) * statistically significant difference from placebo statistically significant heterogeneity #  #  #  Nine of the included trials assessed amlodipine, encompassing a dose range of 1.25 mg/day to 10 mg/day. Amlodipine at 1.25 mg/day did not statistically significandy lower SBP or D B P compared with placebo. The 2.5 mg/day dose is the lowest dose that showed a statistically significant difference compared with placebo in this review.  There was  statistically significant heterogeneity in the 5 mg/day dose ( C h i = 32.56, p < 0.0001) but the 2  random effects model still yielded a statistically significant reduction in SBP and D B P compared with placebo. The heterogeneity can be explained by high baseline BPs (> 160 mm Hg) in two trials (Kuschnir 1996 and Farsang 2001). If these trials are removed, the heterogeneity is no longer statistically significant and the effect size at 5 mg is reduced to 8.78 m m Fig (95% CI -10.80, -6.76) for SBP and -4.10 mm H g (95% CI -5.29, -2.90) for DBP. Indirect comparison of the results for each dose showed evidence of a dose-response phenomenon since there was a greater reduction in blood pressure with 5 mg/day compared with 2.5 mg/day and 1.25 mg/day. Flowever, there was no difference between the 5 and 10 mg/day doses.  150  In the direct comparison analysis, only two trials were included (Frick 1988, Mehta 1993). There was no statistically significant difference in the effect sizes between the direct and indirect comparisons of amlodipine doses. Since all amlodipine trials had a mean age of <60 years and inclusion criteria of elevated D B P , subgroup analyses based on age and type of hypertension were not possible. A l l trials were of moderate quality Qadad score 3 or 4 out of 5) except for one study which had a Jadad score of 5/5 (Pool 2001).  Sensitivity analyses of B P measurement position did  not change the results. In all trials B P measurements were taken in the trough period of the dosing interval; thus an analysis of peak vs. trough effect was not possible.  Only 2  amlodipine trials reported S D of B P change (Mehta 1992 and Mroczek 1988), while imputed values were used for the rest of the trials. The results were insensitive to whether the values were imputed according the hierarchy or directiy from the trials that did report the S D of B P change. From the included studies, the best estimate of the maximal B P lowering efficacy of amlodipine occurring at 5 to 10 mg/day is -10.72 mm H g (95% CI: -12.30, -9.14) for SBP and -5.73 m m H g (95% CI: -6.65, -4.82) for D B P . The log dose-response curve is displayed in Figure 3. A funnel plot of the amlodipine studies at starting dose (5 mg) and above showed an absence of small negative-result trials (Figure 4). Thus, publication bias is likely and the best estimate of the blood pressure lowering efficacy of amlodipine is likely an overestimate of the actual effect size.  151  5n  Log Dose 0.25  I  0.50  Change in S B P Change in DBP  0.75  CL  m <u G) C <o  o  -20  J  Figure 3. L o g dose-response curve for amlodipine, 1.25-10 mg/day.  Review:. Comparison: Outcome:  Blood pressure.lowering efficacy of.caldium channel blockers for primary hypertension 01 Dose Amlodipine v s . placebo 0t change in-Systolic Blood Pressure :  SE(WMD)  +4  -100  -50  50  100  WMD (fixed)  Figure 4. Funnel plot of standard error against effect size of change in SBP for amlodipine 5 to 20 mg/day. Each dot represents an active treatment group.  152  3.7.2 Barnidipine vs. placebo T a b l e 35: B l o o d pressure lowering efficacy of barnidipine 10-30 m g / d a y . Dose of Barnidipine  # of trials  10 mg/day  1  Total # patients in treatment group 50  20 mg/day  1  45  30 mg/day  1  45  Change in SBP (mm Hg)  Change in D B P (mm Hg)  -3.20 (-8.61,2.21) -3.40 (-8.95, 2.15) -5.80* (-11.35, -0.25)  -2.70 (-5.78, 0.38) -3.20 (-6.58, 0.18) -4.90* (-8.06, -1.74)  statistically significant difference from placebo Only 1 trial assessed barnidipine (Hart 1997) and there was no statistically significant difference between all three doses tested but only the 30 mg/day dose decreased B P compared with placebo. The paucity of data is reflected in the wide confidence intervals. N o t enough doses were tested to allow a meaningful dose-response curve to be constructed. The lowest effective dose is 30 mg/day. Based on this one trial, the blood pressure lowering effect across all doses of barnidipine is -4.12 m m H g (95% CI: -7.29, -0.94) for SBP and 3.60 mm H g (95% CI: -5.45, -1.75) for D B P .  Out of the doses tested, the maximal blood  pressure lowering efficacy of barnidipine also occurs at 30 mg/day (-5.80/-4.90 m m Hg) but the true maximal effect cannot be estimated since doses above 30 mg/day have not been tested.  153  3.7.3  Darodipine vs. placebo  T a b l e 36: B l o o d pressure lowering efficacy of darodipine 100-300 m g / d a y . Dose of Darodipine  # of trials  100 mg/day  1  Total # of patients in treatment group 13  200 mg/day  1  9  300 m g / day  1  11  Change in SBP, mm H g (95% CI) -16.70* (-29.05, -4.35) -24.80* (-39.88, -9.72) -15.70* (-28.30, -3.10)  Change in D B P , mm H g (95% CI) -6.70 (-14.48, 1.08) -14.10* (-22.69, -5.51) -11.40* (-19.35, -3.45)  * statistically significant difference from placebo Only 1 trial assessed darodipine (Chrysant 1988) and there was no statistically significant difference between any of the three doses tested.  A l l doses had statistically  significant difference compared with placebo except for change in D B P in the 100 mg/day group. From the available data, the lowest effective dose is 100 mg/day, but it may occur at a lower dose for which there are no data. It is unclear i f the doses tested in this trial reflect the plateau of the dose-response curve or i f doses above 300 mg/day have greater efficacy. Thus the true maximal blood pressure lowering efficacy of darodipine cannot be estimated. However, from the data that are available, it appears that the lowest dose with maximal blood pressure lowering efficacy is 100 mg/day.  Maximum efficacy may be achieved at  doses lower than 100 mg/day but again there are no available data.  The overall blood  pressure lowering effect across all doses of darodipine is -18.40 m m H g (95% CI: -26.02, 10.79) for SBP and -10.51 m m H g (95% CI: -15.17, -5.84) for D B P .  However, since this  result is based on one trial in a small number of individuals, its uncertainty is reflected by the wide confidence limits.  154  3.7.4  Felodipine vs. placebo  T a b l e 37: B l o o d pressure lowering efficacy of felodipine 2.5-20 m g / d a y . Fixed effect model with 95% confidence interval. Change in D B P , Total # of Change in SBP, Dose of # of trials patients in mm H g mm H g Felodipine treatment group (95% CI) (95% CI) . -5.02* 407 -3.16* 2.5 mg/day 5 (-7.25, -2.79) (-4.53,-1.79) -5.12* -5.42* 5 mg/day 6 293 (-8.03, -2.81) (-6.39, -3.86) _7 79* 10 mg/day 335 -10.10* 8 (-11.95, -8.24) (-8.75, -6.83) -6.26* -4.80* 20 mg/day 3 87 (-10.07, -2.44) (-6.70, -2.90) * statistically significant difference from placebo * statistically significant heterogeneity #  #  A l l doses of felodipine resulted in statistically significant reduction in blood pressure compared with placebo. Thus, the lowest effective dose is 2.5 mg/day. However, the true lowest effective dose may be lower than 2.5 mg/day but there are no data available. The 10 mg/day dose had significant inter-study heterogeneity and had statistically significant differences in effect size compared with all other doses tested, even 20 mg/day. However, the result for 20 mg/day is based on only 87 patients, compared with 335 patients in the 10 mg/day category. Differences in baseline B P did not account for this result. The weighted mean baseline B P was 158.3/101.0 mm H g in the 10 mg/day group and 162.7/103.8 m m H g in the 20 mg/day group. The true lowest effective dose may be lower than 10 mg/day but there are no data available Differences in the time of B P measurement relative to dosing appear to account for the superior response observed in the 10 mg/day group.  In 9 out of 11 trials, the B P  measurements were taken at trough, whereas in one trial B P measurements were taken at 2-3 hours post-dose (Kiesewetter 1994) and in another trial the time of measurement was not  155  reported (Fetter 1994). Sensitivity analysis excluding Kiesewetter 1994 and Fetter 1994, both of which tested 10 mg/day vs. placebo, decreased the effect size of this dose [SBP: -7.86 m m H g (-10.24, -5.49); D B P : -5.66 m m H g (-6.90, -4.42)], such that the differences between 5-, 10- and 20 mg/day doses were no longer statistically significant. A s well, the heterogeneity in the 10 mg/day group was no longer statistically significant when these 2 trials were removed. Two felodipine trials had a study population of mean age > 60 years (Black 2001 and van Ree 1996) and one trial reported data separately for older and younger patients (Fagan 1993). Subgroup analysis of studies with patients of mean age > 60 years demonstrated a statistically significandy greater reduction in SBP at 5 mg/day and in D B P at 2.5- and 5 mg/day compared with the younger subgroup. Doses above 5 mg/day were not tested in those trials with mean age of patients > 60 years.  Table 38: Blood pressure lowering efficacy of felodipine in older and younger subgroups Dose of felodipine 2.5 mg/day 5 mg/day  Change i n SBP, mm H g (95% Cf) < 60 years > 60 years -3.61 -6.27 (-6.37, -0.85) (-9.44, -3.10) -11.40 -4.80 (-7.55, -2.06) (-19.95, -2.85)  Change in D B P , m m H g (95% Cf) < 60 years > 60 years -2.44 -5.55 (-9.17,-1.93) (-3.80, -1.09) -4.87 -11.80 (-18.35, -5.25) (-6.15, -3.58)  A l l trials which reported the formulation of felodipine analysed the once-daily extended-release formulation, except the Felodipine Co-op study which tested a twice-daily formulation. Trials with inclusion criteria of elevated SBP (> 140 mg) consisted of Black 2001, Kiesewetter 1994 and van Ree 1996 (Table 39 below).  This subgroup demonstrated a  statistically significandy greater response compared with those trials with inclusion criteria of elevated D B P only. Flowever, this conclusion is based on a small number of patients with  156  elevated SBP. This factor accounted for some but not all of the heterogeneity in the 10 mg/day analysis. T a b l e 3 9 : B l o o d pressure lowering efficacy i n trials w i t h i n c l u s i o n criteria of Dose of felodipine 2.5 mg/day 5 mg/day 10 m g / day  Change in SBP, mm H g (95% CI) Elevated SBP Elevated D B P -5.75 -4.53 (-9.37, -2.14) (-7.40, -1.67) -11.40 -4.80 (-19.95, -2.85) (-7.55, -2.06) -13.00 -8.59 (-16.16, -9.84) (-10.88, -6.30)  Change in D B P , m m H g (95% CI) Elevated SBP Elevated D B P -6.60 -2.99 (-12.79, -0.41) (-4.39, -1.58) -11.80 -4.87 (-18.35, -5.25) (-6.15, -3.58) -11.00 -5.90 (-12.57, -9.43) (-7.10, -4.70)  Sensitivity analyses were performed to assess the robustness of the results.  A  sensitivity analysis removing trials that measured B P in standing or supine positions did not change the results.  Four trials reported the variance of the change in B P , whereas the 7  remaining trials used imputed values. The results were unchanged whether the values were imputed according to the hierarchy or imputed direcdy from the weighted mean values from all trials that reported the S D of B P change. A l l trials were of moderate quality (Jadad score of 3 or 4 out of 5) and thus, a sensitivity analysis based on trial quality was not possible. Blood pressures were measured by sphygmomanometer in all trials that reported the instrument used. A l l trials that reported sources of funding had some funding from pharmaceutical companies involved in the development and marketing of felodipine. Using indirect comparisons, there was no statistically significant difference in trough effect sizes between doses of 2.5-20 mg/day for SBP. For D B P , the 5 mg and 10 mg/day doses each showed a statistically significant difference from the 2.5 mg/day dose, but the magnitude of the differences was of questionaable clinical significance (1.96 m m Fig and 2.5 mm Fig, respectively). Using the subset of multi-arm trials in the direct comparison analysis,  157  the 10 mg/day dose was statistically significantly greater than 2.5 mg/day and 5 mg/day. However, there was no statistically significant difference between the direct and indirect methods for all dose comparisons. In order to include a greater amount of data, the results using the indirect method were used to determine the maximal blood pressure lowering efficacy.  The best estimate of the maximal blood pressure-lowering efficacy at trough for  felodipine 2.5 to 20 mg/day is -6.11 mm H g (-7.41, -4.82) for SBP and -4.75 m m H g (-5.44, 4.05) for D B P . A funnel plot of standard error vs. W M D of the starting dose (5 mg/day) and above demonstrates asymmetry, with an absence of negative-result trials of small sample size (Figure 5).  Thus, publication bias may be present and i f so, the estimate of the blood  pressure lowering efficacy of felodipine is an overestimate of the actual effect size. Revieyy:.' Comparison: Outcome: T  0  "Bipod pressure lowering efficacy of calcium channel blockers' for primary-hypertension 04 Dose Felodipine v s . placebo 01 Change in Systolic Blood Pressure 1  SE(WMD)  •  +3  -100  -50  50  100. WMD (fixed)  Figure 5. Funnel plot of standard error against effect size of change in SBP for felodipine 5 to 20 mg/day. Each dot represents an active treatment group.  158  3.7.5  Isradipine vs. placebo  T a b l e 40: B l o o d pressure lowering efficacy of isradipine 1-20 m g / d a y . Fixed effect model with 95% confidence interval. # of trials Total # of Change in SBP, Change in D B P , Dose of mm H g mm H g Isradipine patients in treatment group (95% CI) (95% CI) 1 . -1.00 1 mg/day 44 -6.00* (-11.35, -0.65) (-4.70, 2.70) -5.22* 2.5 mg/day 4 -8.44* 181 (-7.21, -3.23) (-11.73, -5.15) -10.25* -7.82* 5 mg/day 10 484 (-12.00, -8.50) (-8.85, -6.79) 10 mg/day 4 143 -16.46* -11.55* (-19.75, -13.18) (-13.49, -9.61) 15 mg/day 2 112 -17.00* -10.84* (-20.79, -13.20) (-13.00, -8.69) 20 mg/day 2 107 -15.57* -11.48* (-19.25, -11.89) (-13.80, -9.17) * statistically significant difference from placebo statistically significant heterogeneity #  #  A l l doses of isradipine showed a statistically significant reduction in B P compared with placebo, except for change in D B P in the 1 mg/day group.  Indirect comparisons  showed that the 10-, 15-, and 20 mg/day groups were not significandy different from each other but all were statistically greater than the 1-, 2.5- and 5 mg/day groups. There was statistically significant heterogeneity in the 5 mg/day analysis. A subgroup analysis was performed on drug formulation (twice daily versus once-daily sustained release). The twice-daily formulation at 5 mg showed a statistically significantly greater blood pressure lowering as compared to the sustained-release formulation (Table 41).  159  T a b l e 41: B l o o d pressure lowering efficacy of isradi nne according to formulation Change in D B P , m m H g Dose of Change in SBP, m m H g (95% CI) isradipine (95% CI) . Twice-daily SustainedTwice-daily Sustainedrelease release -8.12 -4.35 2.5 mg -8.77 -6.09 (-13.47,-4.07) (-12.73, -3.51) (-8.91, -3.27) (-7.15,-1.56) -6.55 5mg -12.49 -7.69 -8.98 (-14.85, -10.13) (-10.37, -5.01) (-10.40, -7.56) (-8.03, -5.07) -18.62 -13.45 -10.17 10 mg -15.36 (-16.45, -10.46) (-12.70, -7.63) (-19.98, -10.74) (-23.81, -13.43). 15 mg -15.60 -11.90 -10.20 -20.20 (-12.92, -7.48) (-27.09, -13.31) (-20.64, -10.56) (-15.40, -8.40) -11.80 20 mg -15.70 -15.50 -10.70 (-21.93, -9.47) (-20.40, -10.60) (-15.03, -6.37) (-14.53, -9.07)  The robustness of the effect sizes were tested with sensitivity analyses.  Removing  trials that measured B P at peak or that did not report time of B P measurement reduced the blood pressure lowering effect of isradipine 5 mg/day to -6.99 mm H g (-9.41, -4.57) for SBP and -4.54 m m H g (-5.91, -3.16) for D B P . The analysis did not change the results of other doses, f f the low quality trials (Jadad score < 2) were removed, the effect size of 5 mg/day dose is significandy reduced to -7.24 m m H g (-9.36, -5.12) for SBP and -5.71 mm Pig (-6.92, -4.49) for D B P .  The results at all other doses were not altered with the removal of low  quality trials. The effect size at 5 mg/day was affected by whether the S D of B P change was imputed according to the hierarchy or imputed direcdy from the weighted mean values from all trials that reported the S D of B P change, fn the latter case, the change in SBP was -7.75 m m H g (-9.71, -5.79) and the change in D B P was -5.59 mm H g (-6.72, -4.4).  This  discrepancy is mosdy likely accounted for by the low endpoint variances which were reported in the O'Grady 1997 and Youssef 1992 trials and which were used to impute the S D of change in B P using the hierarchy.  160  Removing trials that did not measure B P in the sitting position or did not report position of measurement only altered the effect size at 5 mg/day, with a change in SBP of 5.21 m m H g (95% C I -8.13, -2.29) and a change in D B P of-4.61 m m H g (95% CI -6.19, 3.03). Only three trials reported funding sources, and all three had industry funding. Only one small trial used an automatic device to measure B P (Arosio 1993), while the rest used the auscultatory method with mercury sphygmomanometers.  Removing this trial did not  alter the results. In one trial (O'Grady 1997) all patients had active atherosclerotic lesions. Again, removing this trial did not change the effect size. The lowest dose with maximal B P lowering occurs at 10 mg/day. A t trough, the best estimate of the maximal blood pressure lowering effect at 10 to 20 mg/day is -14.84 m m H g (-17.47, -12.20) for SBP and -10.57 m m H g (-12.15, -8.98) for D B P . This estimate is based on only one clinical trial (Chrysant 1995); thus, the confidence intervals are wide. Funnel plots of the starting dose (5 mg/day) and above showed there was a possibility of publication bias since there was some asymmetry in the shape of the plot.  161  3.7.6  L a c i d i p i n e vs. placebo  T a b l e 42: B l o o d pressure lowering efficacy of lacidipine 2-4 m g / d a y . Dose of Lacidipine  # of trials  2 mg/day  1  Total # of patients in treatment group 46  4 mg/day  1  39  Change in SBP, mm H g (95% CI) -4.00 (-9.92, 1.92) -7.00* (-13.17,-0.83)  Change in D B P , mm H g (95% CI) -3.10* (-6.10, -0.19) -4.20* (-7.67, -0.73)  * statistically significant difference from placebo  Only 1 included trial assessed lacidipine (Rizzini 1991). The study's population was comprised of elderly hypertensives.  Compared with placebo, the 2 mg/day group had a  statistically significant reduction in D B P but not SBP. The lowest effective dose is 4 mg/day, which showed a statistically significant difference from 2 mg/day. Because there was no data for doses above 4 mg/day, the true maximal blood pressure lowering efficacy cannot be estimated. However, based on the available data, blood pressure lowering occurs at 4 mg/day has a magnitude of -7.00 m m H g (-13.17, -0.83) for SBP and -4.20 mm H g (-7.67, -0.73) for D B P .  162  3.7.7  Lercanidipine vs. placebo  T a b l e 43: B l o o d pressure lowering efficacy of lercanidipine 2.5-20 m g / d a y . Fixed effect model with 95% confidence interval. Change in SBP, Change in D B P , Dose of # of trials Total # of mm H g Lercanidipine patients in mm H g (95% CI) treatment group (95% CI) 1 -1.90 2.5 mg/day 58 -1.20 (-4.89, 1.09) (-5.56, 3.16) 1 -3.60* 5 m g / day 54 -3.80 (-6.58, -0.62) (-8.28, 0.68) 10 mg/day 4 184 -4.31* -11.31** (-13.82, -8.79) (-5.89, -2.72) 1 -4.57* 20 mg/day 9 -9.60 (-7.11, -2.03) (-22.05, 2.85) * statistically significant difference from placebo statistically significant heterogeneity #  The 2.5 mg/day group did not show a statistically significant difference from placebo. There does not appear to be a statistically significant difference between the 10 and 20 mg/day doses in change in SBP, and between 5-20 mg/day in the change in D B P . There was statistically significant heterogeneity in the 10 mg/day group. The 20 mg/day dose did not show a statistically significant difference compared with placebo but this is likely due to the lack of data at this dose. Two trials that assessed lercanidipine 10 mg/day were performed in older patients with high baseline SBP (Barbagallo 2000 and N i n c i 1997). In this subgroup, the change in SBP was -14.55 m m Fig (-17.69, -11.40) and the change in D B P was -3.90 mm H g (-5.85, 1.96). In contrast, the change in SBP in the trials with mean age < 60 years was -5.60 mm H g (-9.74, -1.45), while the change in D B P was similar to that of the older age subgroup at 5.08 m m H g (-7.79, -2.37). One trial selected patients based on elevated SBP (Barbagallo 2000) and did not report the timing of the B P measurement.  Removal of this trial accounted for the  heterogeneity in the 10 mg/day analysis and statistically significantly reduced the change in  163  SBP to -6.60 mm H g (-9.71, -3.49) but did not alter the change in D B P , which was -4.65 m m H g (-6.57, -2.72). One of the included trials was of low quality (Rimoldi 1993). Removal of this trial did not change the results in a statistically significant way. A l l trials that reported details regarding the instrument and timing of measurement had trough BPs taken by auscultation. Only one trial in the analysis measured B P in the sitting position (Ninci 1997). The S D of the B P change was reported in only 1 trial (Omboni 1998) and imputed for the rest of the trials, fmputing using the hierarchy or direcdy from all trials that reported S D of B P change did not alter the results. The best estimate of the maximal blood pressure lowering efficacy of lercanidipine at 10-20 mg/day is -11.24 mm H g (-13.71, -8.78) for SBP and -4.38 m m H g (-5.72, -3.03) for DBP.  164  3.7.8  M a n i d i p i n e vs. placebo  T a b l e 44: B l o o d pressure lowering efficacy of manidipine 10-40 m g / d a y . Fixed effect model with 95% confidence interval. Dose of # of trials Total # of Change in SBP, Change in D B P , mm H g m m Fig Manidipine patients in treatment group (95% CI) (95% CI) 2 -4.89* 10 mg/day 40 -10.37* (-15.31, -5.43) (-7.64, -2.13) 1 20 mg/day -20.10* -11.00* 12 (-30.92, -9.28) (-17.16, -4.84) 40 mg/day 1 13 -23.50* -14.40* (-34.11, -12.89) (-20.43, -8.37) * statistically significant difference from placebo Two of the included trials evaluated manidipine. A l l manidipine doses demonstrated a statistically significant reduction in B P compared with placebo.  One trial (Fogari 1999)  was performed in elderly subjects (mean age of 81.8 years), and had a similar effect size compared with the trial performed in younger patients (Fogari 1996, mean age of 53.4 years). Considering all the doses tested, the lowest effective dose is 10 mg/day. The true lowest effective dose may be lower than 10 mg/day but there are no data available.  Indirect  comparisons showed that the effect sizes of the 20 and 40 mg/day doses were statistically significantly different from that of the 10 mg/day dose. Based on the data from two trials, the maximal blood pressure lowering efficacy of manidipine at 20-40 mg/day is -21.83 mm H g (-29.33, -14.34) and -12.73 m m H g (-17.02, -8.45) for SBP and D B P , respectively; however, because of the small number of patients, this is uncertain.  165  3.7.9  N i c a r d i p i n e vs. placebo  T a b l e 45: B l o o d pressure lowering efficacy of nicardipine 40 - 120 effect model with 95% confidence interval. Total # of Change in SBP, Dose of # of trials mm H g Nicardipine patients in treatment group (95% CI) 1 -19.50* 40 mg/day 15 (-25.26,-13.74) 2 -6.49* 60 mg/day 68 (-11.67, -1.31) 80 mg/day 2 39 -15.37* (-22.97, -7.77) 90 m g / day 2 79 -13.92* (-18.70, -9.13) 2 -13.48* 100 mg/day 38 (-19.46, -7.50) 1 120 mg/day 60 -10.50 * (-15.91,-5.09) * statistically significant difference from placebo  mg/day.  Fixed  Change in D B P , mm H g (95% CI) -7.40* (-13.27,-1.53) -3.86* (-6.50, -1.22) -11.39* (-15.39, -7.39) -6.02* (-8.52, -3.52) -9.21* (-13.75, -4.66) -6.40* (-9.11,-3.69)  Nine of the included trials assessed nicardipine, encompassing a dose range of 40 mg/day to 120 mg/day, but there were few studies at each dose.  A l l doses exhibited a  statistically significandy reduction in SBP and D B P compared with placebo. Considering the data available at the doses tested, the lowest effective dose is 40 mg/day.  In reality the  lowest effective dose may be lower than 40 mg/day but there are no available data. Scuteri 1992 was the only trial that assessed the efficacy of nicardipine at 40 mg/day. The study population consisted of hospitalized patients with mean age 76 years and baseline B P of 171/97. The quality of this trial was low (Jadad score = 2) and the trial did not report the timing of the B P measurements. These reasons may account for the large response at 40 mg/day. Three trials did not report the timing of blood pressure measurement (Asplund 1985, Marcadet 1991, Scuteri 1992). If these trials are removed, there is no statistically significant  166  difference between any of the doses (60 mg/day-120 mg/day), using the indirect comparison method. Two of the eight included nicardipine trials tested a thrice-daily formulation (Soro 1990 and Asplund 1985), while the remainder tested a slow-release twice-daily formulation. Subgroup analysis did not show a statistically significant difference in effect size between the formulations. The effect size was also insensitive to the instrument used for B P measurement and strategy used to impute missing variances. Considering only the trials that reported trough blood pressures, the best estimate of the maximal B P lowering efficacy of nicardipine occurring at 60 to 120 mg/day is -10.49 m m H g (95% CI: -13.20, -7.77) for SBP and -5.98 m m H g (95% Cf: -7.38, -4.57) for D B P .  167  3.7.10 N i f e d i p i n e vs. placebo T a b l e 46: B l o o d pressure lowering efficacy of nifedipine 20 - 100 m g / d a y . Fixed effect model with 95% confidence interval. Total # of Change in SBP, Change in D B P , Dose of # of trials mm H g mm H g Nifedipine patients in (95% CI) treatment group (95% CI) 20 mg/day 143 -5.83 * -5.06* 3 (-8.82, -2.84) (-6.93, -3.18) -9.79 * -6.72* 30 mg/day 4 118 (-8.52, -4.92) (-12.89, -6.69) -29.39** -16.65 40 mg/day 3 57 (-19.90, -13.40) (-35.01, -23.78) 42 50 mg/day 1 -4.40 -4.60* (-10.53, 1.73) (-8.05,-1.15) -12.72* -7.32* 60 mg/day 5 157 (-15.70, -9.74) (-8.95, -5.70) 90 mg/day 1 55 -13.60* -6.80* (-18.45, -8.75) (-9.57, -4.03) 1 100 mg/day 38 -5.80 -6.50* (-12.07, 0.47) (-10.03, -2.97) * statistically significant difference from placebo statistically significant heterogeneity #  #  #  #  #  #  #  #  Twelve of the included trials assessed nifedipine. A l l doses exhibited a statistically significant difference from placebo for change in D B P . A l l doses except 50 mg/day and 100 mg/day showed a statistically significant difference from placebo in the change in SBP. This discrepant result is likely because there was only one trial (Carr 1992) for the 50 and 100 mg/day doses.  The lowest effective dose is 20 mg/day but it may actually be lower;  however, data are not available for doses under 20 mg/day. One trial in the 40 mg/day group (Fadayomi 1986) had an exaggerated effect size of -58.4/-31.7 m m Hg. This trial population consisted of 32 black patients, 12 who were newly diagnosed and the remaining who were inadequately controlled on therapies other than CCBs; the baseline B P was 180/114 mm H g . The publication had reported individual patient data for 30 patients, but blood pressures were measured to the nearest 5 m m H g and the timing of the B P measurement was not reported. Because there is a large reduction in  the standard deviation of SBP from baseline (15.3 mm Hg) to endpoint (8.0 m m Hg) - a pattern that is inconsistent with the pooled data of this systematic review - blinding may have been compromised in this trial and data from this trial must be viewed with suspicion. A sensitivity analysis was performed in which trials with trough measurements were retained and trials that did not report timing of B P measurement were removed (de Simone 1985, Eggertsen 1982, Fadayomi 1986, Ferrera 1984, Serradimigni 1985). In this analysis, there were no trials remaining in the 40 mg/day group, and there was no statistically significant difference between the 30, 60 and 90 mg/day doses.  There was a statistically  significant difference between 20 mg/day and 30 mg/day for reduction in SBP but not for DBP. There was statistically significant heterogeneity in every dose group in the change in D B P . While the heterogeneity in the 20 mg/day group was resolved by removing trials that did not report timing of the B P measurement, the heterogeneity in the other doses remained and could not be explained by any of the predefined sensitivity/subgroup analyses. The best estimate of the maximal blood pressure lowering efficacy at trough for 30 to 100 mg/day is -9.11 mm Fig (-10.69, -7.54) for SBP and -6.14 m m H g (-7.04, -5.23) for DBP.  169  3.7.11. N i l v a d i p i n e vs. placebo T a b l e 47: B l o o d pressure lowering efficacy of nilvadipine 8-30 m g / d a y . Dose of Nilvadipine  # of trials  8 rng/ day  1  Total # of padents in treatment group 60  16 mg/day  1  49  18 mg/day  1  17  Change in SBP, mm H g (95% Cf) -10.90* (-16.58, -5.22) -11.00* (-17.10, -4.90) N o t reported  24 mg/day  1  19  N o t reported  30 mg/day  1  20  N o t reported  Change in D B P , mm H g (95% CI) -5.50* (-8.82,-2.18) -7.80* (-11.88, -3.72) -7.00* (-12.53,-1.47) -6.80* (-12.07,-1.53) -8.20* (-13.40, -3.00)  statistically significant difference from placebo  Only 2 included trials evaluated the blood pressure lowering efficacy of nilvadipine: Hoffmann 1997 tested 8-16 mg/day and Weir 1990 tested 18-30 mg/day.  Although all  nilvadipine groups had a statistsically significant difference from placebo, there was no statistically significant difference between any of the doses assessed.  The  true  maximum  blood pressure lowering efficacy of nilvadipine cannot be estimated since doses below 8 mg/day and above 30 mg/day were not tested. However, based on the available data, the best estimate of the maximal B P lowering efficacy of nilvadipine at trough across all doses tested (8-30 mg/day) is -10.95 m m H g (95% CI: -15.10, -6.79) for SBP and -6.80 m m H g (95% Cf: -8.78, -4.83) for D B P .  This result is based on data from very few patients, and  hence, the confidence intervals are wide.  170  3.7.12 N i s o l d i p i n e vs. placebo T a b l e 48: B l o o d pressure lowering efficacy of nisoldipine 10-30mg/day. Dose of Nisoldipine  # of trials  10 m g / day  1  Total # of patients in treatment group 49  20 mg/day  1  51  30 mg/day  1  48  Change in SBP, mm H g (95% CI) -9.50* (-15.43, -3.57) -15.90* (-21.77, -10.03) -16.20* (-22.09, -10.31)  Change in D B P , mm H g (95% CI) -6.90* (-9.87, -3.93) -9.20* (-12.10, -6.30) -10.60* (-13.51, -7.69)  statistically significant difference from placebo  One included trial assessed nisolidipine at doses of 10, 20 and 30 mg/day (Opie 1997).  A l l doses demonstrated a statistically significant reduction in B P compared with  placebo. Using indirect comparisons there was no statistically significant difference between any of the doses. Based on the results of this one small trial, the best estimate of maximal blood pressure lowering efficacy for nisoldipine 10-30 mg/day is -13.89 m m H g (95% CI: 17.30, -10.49) for SBP and -8.93 m m H g (95 % CI: -10.62, -7.24) for D B P . However, because of the lack of data and doses tested, this result provides very limited information regarding the dose-response of nisolidipine.  171  3.7.13 N i t r e n d i p i n e vs. placebo T a b l e 49: B l o o d pressure lowering efficacy of nitrendipine 5-20 m g / d a y . Fixed effect model with 95% confidence interval. Dose of # of trials Total # of Change in SBP, Change in D B P , Nitrendipine patients in mm H g mm H g treatment group (95% CI) (95% CI) 5 mg/day 1 20 -2.00 -2.00 (-10.69, 6.69) (-7.88, 3.88) 10 mg/day 3 129 -8.34* -5.50* (-11.38, -5.30) (-7.00, -4.00) 20 mg/day 8 154 -13.80* -9.46* (-17.12, -10.49) (-11.54, -7.38) * statistically significant difference from placebo statistically significant heterogeneity #  #  #  #  Nitrendipine at 5 mg/day did not statistically significantiy reduce blood pressure compared with placebo. The lowest effective dose was 10 mg/day. Indirect comparisons did not show a statistically significant difference between the 5 and 10 mg/day groups. The 20 mg/day group demonstrated a statistically significant difference from both the 5 and 10 mg/day groups. There was statistically significant heterogeneity in the 20 mg/day group, most notably in D B P (I =88.9%). One trial had inclusion criteria of isolated systolic hypertension 2  (Paolisso 1991) and a concomitant high baseline SBP (> 170 mm Hg). Removing this trial and another trial with high baseline (Maclean 1990) reduced the heterogeneity in the 20 mg/day group but it was still statistically significant. Six of the ten included studies were of low quality (Jadad score of 2). If these trials are excluded, litde data remain in the analysis but the heterogeneity is no longer statistically significant and the resultant effect size in the 20 mg/day group is -9.17 mm H g (-12.12, -6.21) for SBP and -6.06 m m H g (-7.47, -4.65) for DBP.  172  Two studies had a mean age of participants > 60 years (Gerritsen 1998 and Paolisso 1991) but there were not enough data for a subgroup analysis based on age. The best estimate of the maximal trough blood pressure lowering efficacy at 20 mg/day is -13.80 m m H g (-17.12, -10.49) for SBP and -9.46 (-11.54, -7.38) for D B P . The heterogeneity observed in this dose could only be accounted for by removing low quality studies.  173  3.7.14 Ptanidipine vs. placebo T a b l e 50: B l o o d pressure lowering efficacy of pranidipine 1-8 m g / d a y . Dose of Pranidipine  # of trials  1 mg/day  1  Total # of patients in treatment group 36  2 mg/day  1  37  4 mg/day  1  34  8 mg/day  1  37  Change in SBP, mm H g (95% CI) -4.00 (-12.32, 4.32) -10.00* (-18.31, -1.69) -10.00* (-18.32,-1.68) -9.00* (-17.31, -0.69)  Change in D B P , mm H g (95% CI) -4.00 (-11.06, 3.06) -6.00* (-10.38, -1.62) -10.00* (-15.55, -4.45) -8.00* (-13.55, -2.45)  * statistically significant difference from placebo  One multiarm trial evaluated the blood pressure lowering efficacy of pranidipine (Rosenthal 1996). The mean blood pressures reported in this study were rounded to the nearest m m Fig. The 1 mg/day dose did not exhibit a statistically significant reduction in blood pressure compared with placebo. The lowest effective dose was 2 mg/day. However there was no statistically significant difference between any of the doses tested in change in SBP or D B P . Based on the results of the doses tested in this trial, the best estimate of the maximal trough blood pressure lowering effect of pranidipine at 2 to 8 mg/day is -9.67 mm H g (95% CI: -14.47, -4.87) for SBP and -7.66 (95% CI: -10.59, -4.74) for D B P .  174  3.7.15 Summary of b l o o d pressure lowering efficacy of dihydropyridines T a b l e 51: Summary of b l o o d pressure lowering efficacy of dihydropyridines Dihydropyridine Lowest Lowest dose M a g n i t u d e of m a x i m a l trough B P drug effective with maximal lowering ( m m H g ) , 95% C I dose B P lowering (mg/day) (mg/day)* DBP SBP A #  #  Amlodipine  2.5 mg  Barnidipine  30 mg  Darodipine  fOOmg  Felodipine  2.5 mg  Isradipine  1 mg  Lacidipine  4mg  Lercanidipine  10 mg  Manidipine  fOmg  Nicardipine  40 mg  Nifedipine  20 mg  Nilvadipine  8mg  Nisoldipine  10 mg  Nitrendipine  10 mg  Pranidipine  2mg  5mg  -10.72 -5.73 (-12.30, -9.14) (-6.65, -4.82) 30 mg -5.80 -4.90 (-11.35, -0.25) (-8.06,-1.74) 100 mg -18.40 -10.51 (-26.02, -10.79) (-15.17, -5.84) 2.5 mg -6.11 -4.75 (-7.41, -4.82) (-5.44, -4.05) 10 mg -14.84 -10.57 (-17.47, -12.20) (-12.15, -8.98) 4mg -7.00 -4.20 (-7.67, -0.73) (-13.17, -0.83) 10 mg -11.24 -4.38 (-13.71, -8.78) (-5.72, -3.03) 20 mg -21.83 -12.73 (-17.02, -8.45) (-29.33, -14.34) 60 mg -10.49 -5.98 (-13.20, -7.77) (-7.38, -4.57) 30 mg -9.11 -6.14 (-10.69, -7.54) (-7.04, -5.23) 8mg -10.95 -6.80 (-15.10, -6.79) (-8.78, -4.83) 10 mg -8.93 -13.89 (-10.62, -7.24) (-17.30, -10.49) 20 mg -9.46 -13.80 (-17.12, -10.49) (-11.54, -7.38) 2 mg -9.67 -7.66 (-14.47, -4.87) (-10.59, -4.74) lowest dose for which t lere is a statistically significant difference  * The lowest effective dose is the vs. placebo The lowest dose with maximal blood pressure lowering efficacy is achieved at the dose that exhibits a statistically significantly greater response than all other doses lower than it, whereas doses higher than it do not exhibit a statistically significant difference in effect size. If there was a discrepancy between the determined dose for SBP and DBP, that of SBP is used. A  #  Dihydropyridines were pooled from trials reporting trough blood pressures and were grouped according to increments of the manufactuer's recommended starting dose (Table  175  52). A dose-response is present, with a statistically significant difference between increasing dose increments, except between the two highest dose groups. Thus, maximal B P lowering is achieved at a mean dose of 2.3 times the starting dose and above. The best estimate of the maximal B P lowering for the dihydopyridine subclass is -10.51 m m H g (-11.47, -9.56) for SBP and -6.92 mm H g (-7.42, -6.41) for D B P . T a b l e 52: B l o o d pressure lowering efficacy of dihydropyridines c o m b i n e d according to multiples of starting dose W M D for change W M D for change Dose in SBP, mm H g (expressed as multiple of starting dose, x) in D B P , m m H g (95% CI) (95% CI) Range Mean >0.5x < l x  0.5x  -5.31* (-6.87, -3.74) > l x < 2x 1.1 X -8.26* (-9.14, -7.38) > 2x < 4x -10.13* 2.3 x (-11.25, -9.02) > 4x -11.54* 4.1 x (-13.38, -9.7) statistically significant difference from placebo  -3.35* (-4.32, -5.14* (-5.64, -6.65* (-7.22, -7.81* (-8.86,  -2.38) -4.64) -6.07) -6.76)  3.7.15.1 Dihydropyridines - Analysis of publication bias 3.7.15.1.1. Subgroup analysis based o n trial size A post-hoc analysis was done to determine the relationship between the magnitude of blood pressure lowering and trial size. T o perform a subgroup analysis based on trial size, active treatment arms were categorized into tertiles: the lowest representing the smallest trials, the highest representing the largest trials, and a middle tertile of medium-sized trials. For the sake of consistency, only trials that reported trough B P data were included in this analysis.  For dihydropyridines, comparison of the lowest tertile (n=8-27 patients) and  highest tertiles (n=48-183 patients) demonstrated statistically significant differences in effect  176  size for both SBP and D B P , with larger and more variable effect sizes seen in the smaller trials (Table 53). T a b l e 53: Dihydropyridines: post-hoc subgroup analysis of trough B P lowering based o n trial size Dihydropyridines, Lowest tertile H i g h e s t tertile T r o u g h measurements (8-27 patients) (48-183 patients) Mean of W M D for change in -12.6 mm H g (6.4) -9.0 mm H g (4.1) SBP (SD) # of observations 25 27 p = 0.02 t-test low vs. high tertile Mean of W M D for change in -9.1 m m H g (5.2) -5.4 m m H g (2.2) D B P (SD) # of observations 28 28 t-test low vs. high tertile p = 0.001  T o obtain another estimate of the maximal blood pressure lowering efficacy of dihydropyridines, the tertile analysis was restricted to trials with doses greater than or equal to twice the starting dose. Combined, the middle and highest tertiles give an estimate of -9.1 m m H g (95% CI: -10.8, -7.4) for maximal change in SBP and -6.0 m m H g (95% CI: -5.2, 6.9) for maximal change in D B P . 3.7.15.1.2. T r i m - a n d - f i l l method of adjusting for publication bias The nonparametric ttim-and-fill method was used to adjust the effect size for the presence of publication bias (376). The funnel plot of the maximal blood pressure lowering efficacy of dihydropyridines at greater than or equal to twice the starting dose showed asymmetry, with an absence of negative-result trials (Figures 6 and 7). Data points were filled in to make the funnel plot symmetrical and a new effect size was calculated. For change in SBP, seven data points were filled in and the adjusted effect size was -10.15 mm Fig (-11.08, -9.22). For change in D B P , five data points were filled in and the resulting adjusted effect size was -6.67 mm H g (-7.17, -6.17).  Thus, the original effect size was  overestimated by at least 0.4 mm H g for change in SBP, and by at least 0.3 m m H g for change in D B P . 'Review: ;Blo6d pressure lowering efficacy of calcium channel blockers-for primary^hypertensibn Comparison: 20 Dihydropyridines - analysis based on multiples of starting dose -Outcome: 05 S B P Maximal BP lowering 0  SE(WMD)  +6  •  -100  ^50  50  .100 WMD (fixed)  Figure 6. Funnel plot of change in SBP for dihydropyridines at maximal blood pressure lowering. Each dot represents an active treatment group at >2 times the starting dose.  178  Review;. Blood pressure loweringefficacypt Comparison: 20 Dihydropyridines - analysis based on multiples of starting dose Outcome: 06 DBP Maximal BP lowering -r0  SEiyVMD)  •A•  + 6  -too  :50  100  WMO (fixed)  Figure 7. F u n n e l plot of change i n D B P for dihydropyridines at m a x i m a l b l o o d pressure lowering efficacy.  Each dot represents an active treatment group at >2 times the  starting dose.  179  3.7.16 D i l t i a z e m vs. placebo T a b l e 54: B l o o d pressure lowering efficacy of diltiazem 90 - 540 m g / d a y . Fixed effect model with 95% confidence interval. Change in D B P , Dose of # of trials Total # of Change in SBP, mm H g Diltiazem patients in mm H g treatment group (95% Cf) (95% CI) 1 -0.50 90 mg/day 46 -2.20 (-3.71, 2.71) (-7.80, 3.40) -2.32* -2.91* 120 mg/day 5 329 (-4.85, -0.97) (-3.35,-1.28) -4.12* 236 -2.61* 180 mg/day 3 (-6.54, -1.70) (-3.97,-1.26) -4.04* 240 mg/day 6 224 -6.50* (-8.99, -4.02) (-5.13, -2.95) 300 mg/day 3 92 -7.13* -6.47* (-10.85, -3.41) (-8.49, -4.44) 360 mg/day 4 163 -6.28* -4.96* (-6.55, -3.36) (-9.30, -3.26) 480 mg/day 1 54 -10.80* -6.30* (-15.81,-5.79) (-9.20, -3.40) 1 47 -9.50* -7.90* 540 mg/day (-10.98,-4.82) (-15.35, -3.65) ~ statistically significant difference from placebo * statistically significant heterogeneity #  #  #  y  Twelve of the included trials assessed diltiazem. The 90 mg/day dose did not show a statistically significant difference from placebo. The lowest effective dose of diltiazem was 120 mg/day. Using indirect comparisons between doses, the lowest dose exhibiting maximal blood pressure lowering was determined to be 240 mg/day.  The best estimate of the  maximal blood pressure lowering efficacy of diltiazem for 240 to 540 mg/day is -7.18 m m H g (-8.74, -5.62) for SBP and -5.00 m m Fig (-5.77, -4.23) for D B P . There was statistically significant heterogeneity in the 240 mg/day dose, which was accounted for by removing trials with baseline BPs greater than > 160/100 m m H g (Chan 1997, Djian 1990, Herptn 1990).  180  Two trials were performed in older patients (Chan 1997 and Fiddes 1994).  The  effect sizes were greater in this subgroup compared with younger patients (mean age < 60 years). Although there were not sufficient data for a subgroup analysis, removing these two trials did account for the heterogeneity observed in the 240 mg/day group. A l l trials evaluated sustained-release formulations; in 9 of the trials, dosing was oncedaily, while the other 3 trials used twice-daily regimens. There was no statistically significant difference in effect size between these regimens. A l l trials that reported timing of B P measurement in relation to the dosing interval measured trough values. Seven trials reported funding sources and all of these had industry funding. As well, all trials were of moderate quality and had inclusion criteria of elevated DBP.  Only 2 trials reported BPs measured in the sitting position, while the remaining 10  trials had supine blood pressures. Four of the diltiazem trials reported the S D of B P change, f f trials that did not report the S D of change in B P were imputed direcdy with the weighted mean from all trials that reported this value, the result is not different than using the imputation hierarchy.  181  3.7.17 Verapamil vs. placebo  Table 55: Blood pressure lowering efficacy of verapamil 60 - 540 mg/day. Fixed effect model with 95% confidence interval. Dose of # of trials Total # of Verapamil patients in treatment group 1 60 mg/day 38 100 mg/day  1  53  120 mg/day  6  224  180 mg/day  4  198  200 mg/day  2  119  240 mg/day  7  392  300 mg/day  1  58  360 mg/day  2  156  400 mg/day  1  58  480 mg/day  2  70  540 mg/day  1  55  Change in SBP, mm H g (95% CI) -3.50 (-9.87, 2.87) -1.20 (-6.32, 3.92) -0.95 (-3.55, 1.65) -3.80* (-6.95, -0.65) -5.28* (-9.03,-1.54) -9.22* (-11.47, -6.97) -8.70* (-13.66, -3.74) -10.70* (-14.18, -7.22) -8.80* (-14.34, -3.26) -6.51* (-11.78, -1.24) -17.90* (-23.31, -12.49)  Change in D B P , mm H g (95% CI) -0.70 (-4.03, 2.63) -3.40* (-6.24, -0.56) -1.17 (-2.66, 0.33) -3.29* (-5.02,-1.56) -5.39* (-7.53, -3.25) -5.81* (-7.03, -4.60) -9.40* (-12.09, -6.71) -7.84* (-9.93, -5.76) -8.70* (-11.57,-5.83) -7.09* (-10.00, -4.17) -13.00* (-15.92,-10.08)  * statistically significant difference from placebo Thirteen verapamil studies were included in this systematic review. There were very, few trials at the low-end and high-end of the dose range. The lowest effective dose was 180 mg/day. The lowest dose with maximal B P lowering efficacy is 240 mg/day and the best estimate of the maximal blood pressure lowering efficacy for 240 to 540 mg/day is -9.90 m m H g (-11.43, -8.36) for SBP and -7.46 mm H g (-8.30, -6.61). Sensitivity analyses were performed to assess the robustness of the results. With the exception of two trials that measured B P in supine position, all trials had sitting B P measurements.  Removing the two trials with supine measurements did not change the  182  results. Nine trials reported the variance of the change in B P , whereas the 4 remaining trials used imputed values.  The results were insensitive to whether the values were imputed  according to the hierarchy or imputed direcdy from the weighted mean value from all trials that reported the S D of B P change.  One of the trials had a high baseline B P (Von  Manteuffel 1995) but excluding this trial did not alter the results, nor did removing trials with low quality.  A l l trials tested once-daily sustained-release forms of verpamil.  Subgroup  analysis based on age was not possible since all trials had a mean age of < 60 years. A l l trials had inclusion criteria of elevated diastolic blood pressure and used sphygmomanometers to measure BP. The four trials that reported funding source were all industry-funded.  183  3.7.18 Tiapamil vs. placebo  Dose of Tiapamil  # of trials  300-600 mg/day  1  Total # of patients in treatment group 24  900-1200 mg/day  1  25  Change in SBP, mm H g (95% CI) -4.00 (-12.90, 4.90) -5.00 (-13.47, 3.47)  Change in D B P , mm H g (95% CI) -1.00 (-6.82, 4.82) -4.00 (-9.55, 1.55)  Data from the one included trial that evaluated tiapamil (Blanchett 1991) showed that doses 300-1200 mg/day did not reduce blood pressure compared with placebo. Based on this limited evidence, the lowest effective dose and the maximal blood pressure lowering efficacy of tiapamil cannot be estimated.  184  3.7.19 Summary of b l o o d pressure lowering efficacy of non-dihydropyridines T a b l e 57: Summary of b l o o d pressure NonLowest effective dihydropyridine dose ( m g / d a y ) * * drug  lowering efficacy of non-dihydropyridines Lowest dose M a g n i t u d e of m a x i m a l B P lowering ( m m H g ) with maximal B P lowering DBP SBP (mg/day) -5.00 120 mg 240 mg -7.18 (-8.74, -5.62) (-5.77, -4.23) 240 mg -9.90 -7.46 180 mg (-8.30, -6.61) (-11.43, -8.36) N o doses were statistically significandy different from placebo A #  Diltiazem Verapamil Tiapamil  * The lowest effective dose is the lowest dose for which there is a statistically significant difference vs. placebo The lowest dose with maximal blood pressure lowering efficacy is achieved at the dose that exhibits a statistically significantly greater response than all other doses lower than it, whereas doses higher than it do not exhibit a statistically significant difference in effect size. * If there was a discrepancy between the determined dose for SBP and DBP, that of SBP is used. A  Non-dihydropyridines were analysed as a subclass, according, to increments of the manufacturer's recommended starting dose (Table 58). measurements of B P were included.  Only trials reporting trough  A dose-response is present, with a statistically  significant difference between increasing dose increments, except between the two highest dose groups.  Thus, maximal B P lowering is achieved at a mean dose of 2.3 times the  starting dose and above.  The best estimate of the maximal B P lowering for the non-  dihydopyridine subclass is -8.15 m m H g (-9.50, -6.79) for SBP and -5.97 m m H g (-6.65, 5.28).  185  Table 58: B l o o d pressure lowering efficacy for non-dihydropyridines c o m b i n e d according to multiples of starting dose Dose W M D for change W M D for change in SBP, m m H g in D B P , m m H g (expressed as multiple of starting dose, x) (95% CI) (95% CI) Range Mean >0.5x < l x  -1.18 (-3.32, 0.96) > l x < 2x 1.1 X -4.87* (-5.97, -3.76) > 2x < 4x 2.3 x -7.84* (-9.29, -6.39) > 4x 4.1 x -10.25* (-14.05, -6.45) statistically significant difference from placebo  3.7.19.1  0.6 x  -1.49* (-2.71,-0.26) -3.68* (-4.28, -3.08) -5.84* (-6.56, -5.12) -7.05* (-9.16, -4.94)  Non-dihydropyridines - Assessment of publication bias  Funnel plots of the non-dihydopyridine trials at doses > 2x the starting dose did not show asymmetry. There are not enough data to accurately assess publication bias in this subclass using funnel plots.  3.7.19.1.1 Non-dihydropyridines - subgroup analysis based o n trial size Post-hoc subgroup analysis based on trial sample size showed a trend towards a greater blood pressure lowering effect with larger trials but it was not statistically significant (Table 59).  186  T a b l e 59: N o n - d i h y d t o p y r i d i n e s : post-hoc subgroup analysis of trough b l o o d pressure lowering based o n trial size N o n - dihydropyridines, L o w e s t tertile H i g h e s t tertile trough measurements (15-43 patients) (56-238 patients) Mean of W M D for change in -6.9 mm H g (4.7) -5.8 m m H g (2.4) SBP (SD) # of observations 11 13 t-test lowest vs. highest p = 0.5 tertile Mean of W M D for change in -5.5 mm H g (3.0) -4.7 mm H g (2.4) D B P (SD) # of observations 13 14 t-test lowest vs. highest p = 0.4 tertile  Another estimate of the maximal blood pressure lowering efficacy  of non-  dihydropyridines was obtained by restricting the tertile analysis to trials with doses greater than or equal to twice the starting dose. Combined, the middle and highest tertiles give an estimate of -8.1 m m H g (95% CI: -10.2, -6.1) for maximal change in SBP and -5.7 m m H g (95% CI: -7.3, -4.1) for maximal change in D B P .  187  3.7.20 Other C C B s 3.7.20.1  Lidoflazine vs. placebo  T a b l e 60: B l o o d pressure lowering efficacy of lidoflazine 180 m g / d a y . Dose of # of trials Total # of Change in SBP, Lidoflazine padents in mm H g treatment group (95% Cf) 1 180 mg/day 5 16.00 (-4.34, 36.34)  Change in D B P , mm H g (95% CI) -5.00 (-16.91,6.91)  One trial evaluated lidoflazine in post-infarcdon patients who were not selected on the basis of elevated blood pressure. Since individual patient data were reported, data were extracted for the 10 patients (5 in each of lidoflazine and placebo groups) who met the criteria for hypertension at baseline.  In this trial there was no statistically significant  difference between lidoflazine and placebo in change in blood pressure. Lidoflazine actually exhibited an increase in SBP after 3 months treatment, but this result may be a chance effect in a very small sample.  188  3.7.20.2  M i b e f r a d i l vs. placebo  T a b l e 61: B l o o d pressure lowering efficacy of mibefradil 6.25-100 m g / d a y . Fixed effect model with 95% confidence interval. Change in D B P , Dose of # of trials Total # of Change in SBP, mm H g Mibefradil patients in mm H g (95% CI) treatment group (95% CI) 1 52 -0.20 6.25 mg/day -5.00 (-3.29, 2.89) (-10.56, 0.56) 12.5 mg/day 1 52 -2.40 1.40 (-8.20, 3.40) (-1.46, 4.26) 25 mg/day 1 for SBP 51 for SBP -10.00* -3.58* 2 for D B P 90 for D B P (-5.64, -1.53) (-16.28, -3.72) -5.92* 50 mg/day 1 for SBP 51 for SBP -10.10* 92 for D B P 2 for D B P (-16.44, -3.76) (-8.01, -3.84) 100 mg/day 1 for SBP 52 for SBP -17.00* -10.16* 2 for D B P 92 for D B P (-22.78, -11.22) (-12.33, -7.99) 150 mg/day 0 for SBP 40 N o t reported -12.30* 1 for D B P (-15.05, -9.55) * statistically significant difference from placebo Three mibefradil trials were included in this systematic review. Blood pressure data were not extracted from one of these trials (Oparil 1997) because baseline values were not reported. Another trial (Bernink 1996) only reported D B P data. Doses of 6.25 and 12.5 mg/day did not reduce B P compared with placebo.  Doses 25 mg/day and above had  statistically significant reductions in blood pressure compared with placebo. Thus the lowest effective dose is 25 mg/day. One of the trials (Bursztyn 1997) was performed in elderly patients with a baseline blood pressure of 176/99 m m Fig. There was an insufficient amount of data for a subgroup analysis based on age. Based on the 3 included trials, the maximal blood-pressure-lowering efficacy of mibefradil occurs at 100 mg/day and is best estimated at 100-150 mg/day is -17.00 (-22.78, 11.22) m m H g for SBP (based on one trial) and -10.98 (-12.69, -9.28) for D B P (based on two trials).  189  3.8  B l o o d pressure variability The variability of blood pressure at baseline and/or endpoint was reported for 76  included trials.  The number of observations used in the calculation of the variability  represents the number of active treatment arms from these 76 trials. Seventy-two (95%) of the trials had systo-diastolic hypertension entry criteria, while 3 trials (4%) had isolated systolic hypertension entry criteria.  The remaining trial did not select for hypertensive  patients (Meilink-Hoedemaker 1976). 3.8.1.  Systolic vs. diastolic b l o o d pressure  T a b l e 62: Variability of S B P and D B P at end of treatment SBP  DBP  Weighted mean S D SD of weighted mean S D Weighted mean SBP Weighted mean coefficient of variation (CV)* Number of observations Weighted mean S D S D of weighted mean S D Weighted mean D B P Weighted mean coefficient of variation (CV)* Number of observations t-test: S D of SBP vs. S D of D B P  C C B group 14.5 m m H g 3.9 m m H g 146.6 mm H g 9.8  Placebo group 14.8 m m H g 4.8 m m H g 155.0 m m H g 9.5  73 8.7 m m H g 2.5 mm H g 91.4 m m Pig 9.5  46 8.3 m m H g 2.6 mm Pig 97.5 mm Pig 8.5  79 p < 0.0001  p < 0.0001  p=0.4 t-test: C V SBP vs. C V D B P calculated as weighted mean S D divided by weighted mean B P  49 p =0.06  T o determine whether or not SBP varies to the same degree as D B P , their weighted mean SDs were compared.  As shown in Table 62, the absolute variability of SBP is  statistically significandy greater than that of D B P for both the C C B treatment group and placebo group.  Pfowever, the relative variabilities of SBP and D B P , as expressed by the  coefficient of variation, are similar.  190  3.8.2.  C a l c i u m channel blockers vs. placebo  As shown in Table 62, the weighted mean S D of SBP at the end of treatment was 14.5 m m H g for the C C B group, while the corresponding value for placebo was 14.8 m m H g (p=0.6). For D B P , the weighted mean S D was 8.7 m m H g for C C B group and 8.3 mm Fig for placebo group (p=0.4). Thus, there was no statistically significant difference in B P variability between C C B and placebo treatment. 3.8.3.  Systolic vs. diastolic b l o o d pressure entry criteria  T o determine i f B P entry criteria affected baseline B P variability, comparisons were made between the weighted mean baseline standard deviations of trials that used isolated systolic hypertension as entry criteria, only D B P entry criteria, and trials with criteria of elevated SBP and/or D B P (Table 63).  T a b l e 63: Baseline standard deviations of b l o o d pressure according to entry criteria A l l patients  Trials with only D B P entry criteria (61 trials) 14.5  Trials with both SBP and/or D B P entry criteria* (11 trials) 13.3  Trials with only SBP entry criteria (3 trials) 7.9  Weighted mean S D of SBP at baseline (mm Hg) 3.2 3.7 3.6 SD of weighted mean S D of SBP (mm Hg) 28 Number of 165 6 observations t-test systo-diastolic hypertension vs. isolated systolic hypertension: p < 0.0001  5.3 Weighted mean S D 5.5 6.1 of D B P at baseline (mm Hg) SD of weighted 1.6 2.5 2.0 mean S D of D B P (mm Hg) Number of 176 6 28 observations t-test systo-diastolic hypertension vs. isolated s ystolic hypertension: p = 0.4 * includes those trials using World Health Organization classification of hypertension :  191  Trials including patients  with isolated  systolic  hypertension  had statistically  significantly lower baseline SBP variability than trials that based their entry criteria on elevated D B P or mixed S B P / D B P criteria. However there was no statistically significant difference in the variability of D B P among all three types of entry criteria.  3.8.4.  Baseline vs. endpoint  For trials with D B P entry criteria, the standard deviations of B P at baseline and endpoint were compared in Table 64.  There was no statistically significant difference  between the variability of SBP at baseline and endpoint for both the CCB-treated groups and the placebo groups. For variability of D B P , the baseline SDs were statistically significandy lower than the endpoint values for both the C C B and placebo groups.  Table 64: Standard deviations of BP at baseline vs. endpoint in trials with D B P entry criteria C C B group Placebo group 14.6 (3.2) 13.9 (3.3) Weighted A t baseline (SD) mean S D of A t endpoint (SD) 14.9 (4.8) 14.6 (3.8) SBP p = 0.2 t-test baseline vs. p = 1.0 endpoint Weighted mean S D of DBP  A t baseline (SD) A t endpoint (SD) t-test baseline vs. endpoint  5.5 (2.1) 8.7 (2.5) p < 0.0001  5.3 (1.9) 8.3 (2.3) p < 0.0001  192  3.9  Pulse Pressure Although none of the included trials reported pulse pressure as a primary or  secondary outcome, the change in pulse pressure was calculated by subtracting the change in D B P from the change in SBP for each trial that reported both parameters. A weighted mean change in pulse pressure was then calculated, along with the associated standard deviation. Table 65 below shows the change in pulse pressure at trough for dihydropyridines and nondihydropyridines at twice the starting dose and above (i.e., at maximal B P lowering). The effect size of placebo was calculated from all trials reporting both SBP and D B P data. Table 65: Change in pulse pressure # of studies Dihydropyridines  43  Weighted mean change in pulse pressure (95% CI) -3.4 (-4.3, -2.5)  Non-dihydropyridines  18  -2.4 (-3.7,-1.1)  Placebo  94  0.3 (-0.2, 0.8)  Dihydropyridines vs. non-dihydropyridines: p=0.2, N S Dihydropyridines vs. placebo: p < 0.001 Non-dihydropyridines vs. placebo: p < 0.001  Placebo treatment does not have a statistically significant effect on change in pulse pressure, but both the dihydropyridines and non-dihydropyridines demonstrated statistically significant reductions in pulse pressure. Because there is no statistically significant difference between dihydropyridines and non-dihydropyridines, the results of both subclasses were combined to give an overall change in pulse pressure of -3.1 m m H g (95% CI: -3.8, -2.3).  193  3.10  Dose-related change i n heart rate of individual C C B drugs  Forty-seven of the 106 included trials (44%) had heart rate data that either were extractable from the publications or were obtained dtrectiy from authors.  The results are  described for each individual drug as well as grouped into dmydropyridines and nondihydropyridines, categorized according to increments of the manufacturer's recommended starting dose. Lidoflazine and mibefradil are considered separately.  3.10.1 Dihydropyridines vs. placebo Thirty-eight dihydropyridine trials reported heart rate data (Table 66). In the group of trials at 1-1.33 times the starting dose, there was no statistically significant difference in heart rate compared with placebo. There was a statistically significant increase in heart rate in the middle dose range of 1.5-3 times the starting dose.  There was also statistically  significant heterogeneity in this group, stemming from one trial (Rimoldi 1994) that had a very large increase in heart rate in the treatment group. The highest dosage range of > 3 times the starting dose had a non-statistically-significant increase in heart rate compared with placebo; however only 8 studies were included in this category and thus the analysis lacked power. A l l trials reported trough changes in heart rate except for two trials (Kiesewetter 1994, Prisant 1991), which reported peak values. Sensitivity analysis removing these 2 trials did not alter the results.  194  T a b l e 66: Effect of dihydropyridines o n heart rate Dose (expressed dose, x) Range  as multiple of starting  Number trials  of  W M D - all trials (95% confidence interval)  Mean  lx-1.33x  l.Ox  20  1.5x-3x  2.2x  21  > 3x  4.1x  8  0.60 (-0.27, 1.47) 1.43* (0.52, 2.33) 1.32 (-0.46, 3.09) #  * statistically significant difference from placebo statistically significant heterogeneity #  3.10.1.1  A m l o d i p i n e vs. placebo  T a b l e 67: Effect of amlodipine o n heart rate in Dose of Amlodipine # of trials Total number of Change patients in treatment beats/minute group • (95%CI) -2.20 1.25 mg/day 1 47 (-5.73, 1.33) 2.5 mg/day 1 45 -1.30 (-4.87, 2.27) 5 mg/day 3 168 0.28 (-1.58, 2.13) 1 15 1.60 10 mg/day (-5.06, 8.26)  HR,  Based on 4 trials that reported heart rate data (Frick 1988, Mroczek 1988, Pool 2001, Licata 1993), none of the doses of amlodipine showed a statistically significant change in heart rate compared with placebo.  195  3.10.1.2  Darodipine vs. placebo  T a b l e 68: Effect of darodipine o n heart rate Dose of Darodipine # of trials Total number of patients in treatment group 100 mg/day 1 13 200 mg/day  1  9  300 mg/day  1  11  Change in beats/minute (95% CI) -10.70* (-19.15, -2.25) 10.20 (-0.35, 20.75) 3.20 (-4.99, 11.39)  HR,  statistically significant difference from placebo The one included trial that evaluated darodipine reported change in heart rate. Darodipine at 100 mg/day showed a statistically significant decrease in heart rate compared with placebo, while the 200 mg/day and 300 mg/day doses demonstrated non-statisticallysignificant increases in heart rate. 3.10.1.3  Felodipine vs. placebo  T a b l e 69: Effect of felodipine o n heart rate Dose of Felodipine  # of trials  5 mg/day  1  Total number of patients in treatment group 27  10 mg/day  3  123  20 mg/day  2  51  Change in beats/minute (95% CI) 1.50 (-5.87, 8.87) 1.48 (-0.57, 3.52) 1.45 (-1.34, 4.23)  HR,  Three felodipine trials reported change in heart rate (Felodipine Co-op, Liedholm 1989, Kiesewetter 1994). None of the doses showed a statistically significant difference in heart rate compare with placebo.  196  3.10.1.4  Isradipine vs. placebo  T a b l e 70: Effect of isradipine o n heart rate Dose of Isradipine  # of trials  1 nig/day  1  Total number of patients in treatment group 44  2.5 mg/day  2  105  5 mg/day  3  138  10 mg/day  2  45  15 mg/day  1  34 >  20 mg/day  1  30  Change in beats/minute (95% CI) 1.00 (-2.66, 4.66) 1.37 (-0.90, 3.64) 0.40 (-1.74, 2.54) 3.90 (-0.03, 7.83) 6.00* (0.75, 11.25) 3.40 (-1.99, 8.79)  HR,  * statistically significant difference from placebo Four isradipine trials reported change in heart rate (Burger 1993, Italian-Belgian, Homes 1993, Prisant 1991).  None of the doses had a statistically significant difference  compared with placebo, except for the 15 mg/day dose, which was based on 1 trial.  3.10.1.5  Lercanidipine vs. placebo  T a b l e 71: Effect of lercanidpine o n heart rate  10 mg/day  4  Total number of patients in treatment group 176  20 mg/day  3  139  Dose Lercanidipine  of  # of trials  Change in beats/ minute (95% CI) 0.87 (-0.80, 2.54) 1.25 (-0.97, 3.46)  HR,  Five lercanidipine trials reported heart rate data (Barbagallo 2000, Circo 1997, Ninci 1997, Rimoldi 1993 and Rimoldi 1994).  There was no statistically significant difference  between lercanidipine at 10-20 mg/day and placebo in change from baseline in heart rate.  197  3.10.1.6  M a n i d i p i n e vs. placebo  T a b l e 72: Effect of manidipine o n heart rate Dose of Manidipine  # of trials  10 mg/day  2  Total number of patients in treatment group 40  20 mg/day  1  12  40 m g / day  1  13  Change in beats/minute (95% CI) 1.46 (-1.98, 4.90) 1.00 (-8.62, 10.62) 2.00 (-7.42, 11.52)  HR,  Two trials reported heart rate data for manidipine (Fogari 1996 and Fogari 1999). There was no significant difference in change in heart rate between placebo and manidipine at 10-40 mg/day. 3.10.1.7  N i c a r d i p i n e vs. placebo  T a b l e 73: Effect of nicardipine o n heart rate Dose of Nicardipine  # of trials  40 mg/day  1  Total number of patients in treatment group 15  60 mg/day  1  11  80 mg/day  2  39  90 mg/day  1  24  100 mg/day  2  38  Change in beats/minute (95% CI) -2.20 (-7.90, 3.50) -0.80 (-8.75, 7.15) -0.13 (-3.80, 3.55) 2.00 (-3.49, 7.49) -0.14 (-4.28, 4.00)  HR,  Seven trials reported heart rate data for nicardipine (Asplund 1985, Bellet 1987, D e Cesaris 1993, Marcadet 1991, Mazzola 1988, Scuten 1992, Soro 1990). A t all doses  198  tested (40-100 mg/day), nicardipine was not associated with a statistically significant change in heart rate compared with placebo. 3.10.1.8  N i f e d i p i n e vs. placebo  T a b l e 74: Effect of nifedipine o n heart rate Dose of Nifedipine  # of trials  20 mg/day  2  Total number of patients in treatment group 103  30 mg/day  1  62  40 mg/day  2  42  60 mg/day  3  112  90 mg/day  1  59  Change in beats/minute (95% CI) -0.61 (-1.77,2.99) -2.50 (-5.81, 0.81) -0.34 (-4.63, 3.96) -1.50 (-4.09, 1.09) -1.40 (-4.71, 1.91)  HR,  Seven trials reported heart rate data for nifedipine (DeSimone 1985, Feig 1993, Ferrera 1984, Harder 1994, Jueng 1987, Serradimigni 1985, Toal 1997).  Doses 20-90  mg/day did not exhibit statistically significant differences in heart rate compared with placebo. 3.10.1.9  N i s o l d i p i n e vs. placebo  T a b l e 75: Effect of nisoldipine o n heart rate Dose of Nisoldipine  # of trials  10 mg/day  1  Total number of patients in treatment group 49  20 mg/day  1  51  30 mg/day  1  48  Change in beats/minute (95% CI) 2.90 (-1.02, 6.82) 2.20 (-1.71,6.11) 5.00* (1.06, 8.94)  HR,  * statistically significant difference from placebo  199  One trial reported heart rate data for nisoldipine (Opie 1997). A l l doses showed an increase in heart rate but only the effect size at 30 mg/day was statistically significant compared with placebo. 3.10.1.10  N i t r e n d i p i n e vs. placebo  T a b l e 76: Effect of nitrendipine o n heart rate Dose of nitrendipine  # of trials  10 mg/day  1  Total number of patients in treatment group 72  20 mg/day  3  47  Change in beats/minute (95% CI) 2.70* (0.65, 4.75) 4.70* (0.35, 9.05)  HR,  statistically significant difference from placebo Four trials reported heart rate data for nitrendipine (Asmar 1992, Lederle 1991, Ferrera 1985, Paolisso 1991). Both the 10 mg and 20 mg/day groups showed statistically significant increases in heart rate compared with placebo.  3.10.2 Non-dihydropyridines vs. placebo Seven non-dihydropyridine studies reported heart rate data (Table 77). While both the low and middle dose categories showed decreases in heart rate, only the middle dose range achieved statistical significance. N o trials reported heart rate data for doses above 3 times the starting dose.  200  T a b l e 77: Effect of non-dihydropyridines (diltizaem and verapamil) o n heart rate Dose (expressed dose, x) Range  as multiple of starting  Number trials  of  WMD (95% interval)  confidence  Mean  lx-1.33x  l.lx  5  1.5x-3x  2.1x  4  > 3x n/a 0 * statistically significant difference from placebo  3.10.2.1  -1.49 (-3.23, 0.25) -3.76* (-5.53, -1.99) n/a  D i l t i a z e m vs. placebo  T a b l e 78: Effect of diltiazem on heart rate Dose of Diltiazem  # of trials  240 mg/day  1  Total number of patients in treatment group 24  300 mg/day  2  80  360 mg/day  1  23  Change in beats/minute (95% CI) 1.20 (-6.12, 8.52) -3.41* (-6.35, -0.47) 0.90 (-5.77, 7.57)  HR,  * statistically significant difference from placebo Two trials reported heart rate data for diltiazem (Djian 1990, Meeves 1994). Based on these limited data there was a statistically significant decrease in heart rate in the 300 mg/day analysis but non-significant increases in heart rate for the 240 mg/day and 360 mg/day doses.  201  3.10.2.2  Verapamil vs. placebo  Table 79: Effect of verapamil o n heart rate  Dose of Verapamil  # of trials  120 mg/day  3  Total number of padents in treatment group 117  180 mg/day  3  141  240 mg/day  1  27  360 mg/day  1  59  480 mg/day  1  26  540 mg/day  1  55  Change in beats/minute (95% CI) -2.44* (-4.59, -0.28) -1.22 (-3.38, 0.94) -2.60 (-5.81, 0.61) -5.30* (-8.22, -2,38) -1.00 (-5.39, 3.39) -5.80* (-8.86, -2.74)  HR,  statistically significant difference from placebo Four trials (Carr 1991, Scholze 1998, V E R A T R A N , White 1995) reported heart rate data for verapamil over a dose range of 120 — 540 mg/day. The 120 mg/day, 360 mg/day and 540 mg/day groups showed a statistically significant decrease in heart rate compared with placebo, while the 180 mg/day, 240 mg/day, and 480 mg/day groups did not. However, the results for 240-540 mg/day are based on only one trial at each dose.  3.10.2.3  T i a p a m i l vs. placebo  Table 80: Effect of tiapamil o n heart rate  Dose of tiapamil  # of trials  300-600 mg/day  1  Total number of patients i n treatment group 24  900-1200 mg/day  1  25  Change in beats/minute (95% CI) 3.00 (-2.58, 8.58) 2.00 (-4.02, 8.02)  HR,  202  The effect of tiapamil on heart rate was examined in one included trial (Blanchett 1991).  There was no statistically significant difference in heart rate at doses 300-1200  mg/day compared with placebo.  3.10.3 Other c a l c i u m channel blockers 3.10.3.1  Lidoflazine vs. placebo  T a b l e 81: Effect of lidoflazine o n heart rate Dose of lidoflazine  # of trials  180 mg/day  1  Total number of patients in treatment group 5  Change in beats/minute (95% CI) -15.20* (-24.99, -5.41)  HR,  *statistically significant difference from placebo The one included trial assessing lidoflazine (Meilink-Hoedemaker 1976) reported heart rate data. There was a statistically significant drop in heart rate with lidoflazine 180 mg/day compared with placebo. 3.10.3.2  M i b e f r a d i l vs. placebo  T a b l e 82: Effect of mibefradil o n heart rate Dose of mibefradil  # of trials  6.25 mg/day  1  Total number of patients in treatment group 52  12.5 mg/day  1  52  25 mg/day  1  51  50 mg/day  1  51  100 mg/day  1  52  Change in beats/minute (95% CI) -0.80 (-4.17, 2.57) -0.40 (-4.03, 3.23) -2.00 (-5.53, 1.53) -5.20* (-9.01,-1.39) -6.60* (-10.13, -3.07)  HR,  * statistically significant difference from placebo  203  One included trial (Bursztyn 1997) reported heart rate data for mibefradil. The 50 and 100 mg/day doses had a statistically significant decrease in heart rate compared with placebo, while the effect sizes at lower doses were not statistically significant.  204  3.11  Dose-related withdrawals due to adverse events  Analysis of withdrawals due to adverse events during 3 to 12 weeks of treatment was based on 64/106 (60%) of the included studies (50 dihydropyridine trials and 14 nondihydropyridine trials).  3.11.1 Dihydropyridines vs. placebo O f patients receiving a dihydropyridine, 196/3944 (5%) withdrew due to adverse events, compared with 60/2062 (3%) patients treated with placebo.  The pooled result  demonstrated a statistically significant difference (RR 1.49 [95% CI 1.12, 1.98]), without heterogeneity. Whether each dihydropyridine is considered independendy or grouped together as a subclass, there appears to be a trend of increased W D A E with increase dose. Possible doseresponse relationships in the chhydropyridine subclass were evaluated by combining trials in increments of the recommended starting doses (Table 83).  T a b l e 83: Effect of dihydropyridines on withdrawals due to adverse events Dose (expressed as multiple of starting dose, x) Range Mean  # of trials  # of W D A E in treatment group vs. placebo group  tx-1.33x l.Ox 30 47/1570 vs. 45/1397 1.5x-3x 2.0x 31 80/1388 vs 33/1098 4.1x 11 53/423 vs 10/392 > 3x statistically significant difference from placebo  RR (95% CI)  0.97 (0.66, 1.43) 1.76 (1.21,2.58)* 3.91 (2.19, 6.99)*  205  3.11.1.1  Amlodipine vs. placebo  Table 84: Effect of amlodipine on withdrawals due to adverse events Dose of Amlodipine  # of trials  1.25 mg 2.5 mg  1 2  # of W D A E in treatment group vs. placebo group 0/41 vs. 0/40 5/124 vs. 1/122  5mg  4  10/501 vs. 12/381  10 mg  1  1/43 vs. 0/40  RR (95% CI) Not estimable 3.61 (0.61, 21.58) 0.75 (0.34, 1.67) 2.80 (0.12, 66.70)  Six trials reported W D A E data for amlodipine (Chrysant 2003, Farsang 2001, Frishman 1995, Kuschnir 1996, Mehta 1993, Pool 2001). None of the results showed a statistically significant difference from placebo.  3.11.1.2  Darodipine vs. placebo  Table 85: Effect of darodipine on withdrawals due to adverse events Dose of Darodipine  # of trials  100 mg/day 200 mg/day 300 mg/day  1 1 1  # of W D A E in treatment group vs. placebo group 0/13 vs. 0/10 0/9 vs. 0/10 0/11 vs. 0/10  RR (95% CI) Not estimable Not estimable Not estimable  One darodipine trial (Chrysant 1988) reported W D A E . Since no withdrawals due to adverse events occurred in this trial, this outcome cannot be assessed in this review.  206  3.11.1.3  Felodipine vs. placebo  Table 86: Effect of felodipine on withdrawals due to adverse events Dose of Felodipine  # of trials  2.5 mg/day  2  # of W D A E in treatment group vs. placebo group 3/100 vs. 3/99  5 mg/day  5  9/221 vs. 6/214  10 mg/day  6  19/289 vs. 9/288  20 mg/day  3  30/118 vs. 2/123  RR (95% CI) 0.99 (0.20, 4.79) 1.35 (0.53, 3.42) 2.01 (0.95, 4.23) 10.58* (3.32, 33.66)  statistically significant difference from placebo Seven trials reported W D A E data for felodipine (Felodipine Co-op, Liedholm 1989, Kiesewetter 1994, Scholze 1999, van Ree 1996, Weber 1994, Wester 1991). There appears to be a dose-related increase in W D A E and the 20 mg/day group demonstrates a statistically significant increase compared with placebo.  With all doses pooled, there is a statistically  significant relative risk of 2.75 (95% CI 1.73, .4.36).  3.11.1.4  Isradipine vs. placebo  Table 87: Effect of isradipine on withdrawals due to adverse events Dose of 1 sradipine  # of trials  1 mg/day  1  # of W D A E i n treatment group vs. placebo group 0/45 vs. 1/44  2.5 mg/day  5  6/187 vs. 3/154  5 mg/day  6  9/353 vs. 6/263  10 mg/day  2  2/38 vs. 0/34  RR (95% CI) 0.33 (0.01, 1.48 (0.47, 1.06 (0.38, 4.78 (0.24,  7.80) 4.64) 2.96) 94.12)  207  Eight trials reported W D A E data for isradipine (Bellet 1987, Burger 1993, Holmes 1993, Italian-Belgian, Kirch 1990, M a n in't Veld 1991, O'Grady 1997, Pittrow 1997). There were few trials reporting data for each dose, and thus, due to limited power, none of the results showed a statistically significant difference from placebo.  3.11.1.5  Lercanidipine vs. placebo  Table 88: Effect of lercanidipine on withdrawals due to adverse events Dose Lercanidipine  of  # of trials  2.5 mg/day  1  # of W D A E in treatment group vs. placebo group 2/63 vs. 3/60  5 mg/day  1  0/59 vs. 3/60  10 mg/day  3  4/113 vs. 5/111  20 mg/day  1  1/10 vs. 0/10  RR (95% CI) 0.63 (0.11,3.67) 0.15 (0.01, 2.75) 0.78 (0.22, 2.84) 3.00 (0.14, 65.90)  Three studies reported W D A E data for lercanidipine (Barbagallo 2000, O m b o n i 1988, Rimoldi 1994). There were few trials reporting data for each dose, and thus, due to limited power, none of the results showed a statistically significant difference from placebo.  3.11.1.6  Manidipine vs. placebo  Table 89: Effect of manidipine on withdrawals due to adverse events Dose of Manidipine  # of trials  10 mg/day 20 mg/day  2 1  # of W D A E in treatment group vs. placebo group 0/40 vs. 0/40 1/13 vs. 0/13  40 mg/day  1  0/13 vs. 0/13  RR (95% CI) Not estimable 3.00 (0.13, 67.51) N o t estimable  208  Two trials reported W D A E data for manidipine (Fogari 1996 and Fogari 1999). There were insufficient data to estimate the effect of manidipine on W D A E .  3.11.1.7  N i c a r d i p i n e vs. placebo  T a b l e 90: Effect of nicardipine on withdrawals due to adverse events Dose of Nicardipine  # of trials  60 mg/day  2  # of W D A E in treatment group vs. placebo group 7/78 vs. 3/69  80 mg/day 90 mg/day  1 2  0/15 vs. 0/15 6/90 vs. 3/87  100 mg/day 120 mg/day  2 1  0/39 vs. 0/39 12/68 vs. 3/63  RR (95% CI) 1.86 (0.54, 6.35) Not estimable 1.97 (0.51, 7.53) N o t estimable 3.71* (1.10, 12.53)  statistically significant difference from placebo Six trials reported W D A E data for nicardipine (Asplund 1985, Bellet 1987, Fagan 1993, Marcadet 1991, Mazzola 1988, Soro 1990). Based on the result of one trial (Fagan 1993,  there was a statistically significant increase in W D A E in the 120 mg/day group  compared with placebo.  When all doses were pooled, there was a statistically significant  relative risk of 2.48 (95% CI: 1.21, 5.08).  209  3.11.1.8  N i f e d i p i n e vs. placebo  T a b l e 91: Effect of nifedipine o n withdrawals due to adverse events Dose of Nifedipine  # of trials  20 m g / day  2  # of W D A E in treatment group vs. placebo group 3/143 vs. 7/145  30 mg/day  3  8/116 vs. 2/113  40 mg/day  3  5/79 vs. 5/79  50 mg/day  1  4/53 vs. 4/51  60 mg/day  5  7/166 vs. 3/167  90 mg/day  1  7/59 vs. 0/62  100 mg/day  1  6/53 vs. 4/51  RR (95% CI) 0.47 (0.14, 1.64) 3.35 (0.83, 13.55) 1.25 (0.35, 4.44) 0.96 (0.25, 3.64) 2.01 (0.62, 6.54) 15.75 (0.92, 269.79) 1.44 (0.43, 4.82)  Ten trials reported W D A E data for nifedipine over a range of 20-100 mg/day (Carr 1992, Eggersen 1982, Fadayomi 1986, Feig 1993, Harder 1994, Jueng 1987, Serradimigni 1985, Toal 1997, Zachariah 1990, Zanchetti 1994).  There were few trials reporting data for  each dose, and thus, due to limited power, none of the doses showed a statistically significant difference from placebo. The pooled result across all doses showed a marginally statistically significant increase in W D A E compared with placebo (RR 1.60, 95% CI: 1.00, 2.56).  3.11.1.9  N i l v a d i p i n e vs. placebo  T a b l e 92: Effect of nilvadipine o n withdrawals due to adverse events Dose of Nilvadipine  # of trials  8 mg/day  1  # of W D A E in treatment group vs. placebo group 6/61 vs. 2/59  16 mg/day  1  7/52 vs. 2/59  RR (95% CI) 2.90 (0.61,13.81) 3.97 (0.86, 18.28)  210  One trial (Hoffman 1997) reported W D A E data for nilvadipine. Due to limited power, none of the doses showed a statistically significant difference from placebo. Pooled results did show a statistically significant increase in W D A E compared with placebo, with a relative risk of 3.41 (95% CI: 1.15, 10.13).  3.11.1.10  Nisoldipine vs. placebo  Table 93: Effect of nisoldipine on withdrawals due to adverse events Dose of Nisoldipine  # of trials  10 mg/day  1  # of W D A E in treatment group vs. placebo group 2/49 vs. 1/58  20 mg/day  1  0/51 vs. 1/58  30 mg/day  1  0/48 vs. 1/58  RR (95% CI) 2.37 (0.22, 25.33) 0.38 (0.02, 9.08) 0.40 (0.02, 9.63)  One trial (Opie 1997) reported W D A E data for nisoldipine. Due to limited power, none of the results showed a statistically significant difference from placebo.  3.11.1.11  Nitrendipine vs. placebo  Table 94: Effect of nitrendipine on withdrawals due to adverse events Dose Nitrendipine  of  # of trials  10 mg/day  2 .  # of W D A E in treatment group vs. placebo group 4/116 vs. 3/105  20 mg/day  3  2/61 vs. 0/53  RR (95% CI) 1.17 (0.27, 5.07) 5.00 (0.25, 100.20)  211  Five trials reported W D A E data for nitrendipine (Asmar 1993, Ferrara 1985, Fodor 1991, Lederle 1991, Maclean 1990). Due to limited power, none of the results showed a statistically significant difference from placebo.  3.11.1.12  Pranidipine vs. placebo  T a b l e 95: Effect of pranidipine on withdrawals due to adverse events Dose of Pranidipine  # of trials  1 mg/day  1  # of W D A E in treatment group vs. placebo group 0/36 vs. 3/32  2 mg/day  1  0/37 vs. 3/32  4 mg/day  1  3/34 vs. 3/32  8 mg/day  1  5/37 vs. 3/32  RR (95% CI) 0.13 (0.01,2.38) 0.12 (0.01, 2.31) 0.94 (0.20, 4.33) 1.44 (0,37, 5.57)  One trial (Rosenthal 1996) reported W D A E for pranidipine. Although there was a trend towards increased W D A E with higher doses, none of the doses had a statistically significant difference from placebo due to limited power.  3.11.2 Non-dihydropyridines vs. placebo O f patients receiving a non-dihydropyridine (diltiazem or verapamil), 31/1258 (2.5%) withdrew due to adverse events, compared with 16/599 (2.7%) treated with placebo. There was no statistically significant difference between active treatment and placebo in the pooled result (RR 0.84 [95% CI 0.51, 1.38]). There does not appear to be a dose-response relationship for W D A E in diltiazem and verapamil trials but there are too few trials to make an accurate estimation of the effect size for this outcome (Table 96).  212  T a b l e 96: Effect of non-dihydropyridines o n withdrawals due to adverse events Dose (expressed as multiple of starting dose, x) Range Mean  # of trials  # of W D A E in treatment group vs. placebo group  RR (95% CI)  lx-1.33x 1.5x-3x > 3x  9 7 2  12/558 vs. 16/543 17/598 vs. 3/428 2/102 vs. 4/99  0.75 (0.37,1.54) 1.08 (0.5, 2.31) 0.47 (0.08, 2.64)  l.lx 2.2x 4.3x  3.11.2.1  D i l t i a z e m vs. placebo  T a b l e 97: Effect of diltiazem o n withdrawals due to adverse events Dose of Diltiazem  # of trials  90 mg/day  1  # of W D A E in treatment group vs. placebo group 0/46 vs. 2/43  120 mg/day  3  5/229 vs. 8/233  180 mg/day  2  9/198 vs. 8/193  240 mg/day  2  0/80 vs. 2/83  300 mg/day 360 mg/day  1 2  0/55 vs. 0/56 6/103 vs. 4/99  480 mg/day  1  1/55 vs. 2/56  540 mg/day  1  1/47 vs. 2/43  RR (95% CI) 0.19 (0.01, 3.79) 0.65 (0.23, 1.88) 1.10 (0.43, 2.79) 0.21 (0.01, 4.22) Not estimable 1.43 (0.42, 4.93) 0.51 (0.05, 5.45) 0.46 (0.04, 4.87)  Six trials reported W D A E for diltiazem (Levine 1995, McMahon 1989, Neutel 1996, Neutel 1999, Scholze 1998, Smith 2001). There was not enough power to achieve statistical significance because of the limited amount of data.  213  3.11.2.2  V e r a p a m i l vs. placebo  T a b l e 98: Effect of verapamil o n withdrawals due to adverse events Dose of Verapamil  # of trials  # of W D A E in treatment group vs. placebo group 1/42 vs. 2/45  60 mg/day  1  100 mg/day 120 mg/day  1  3  0/53 vs. 0/51 0/104 vs. 3/104  180 mg/day  1  0/29 vs. 1/30  200 mg/day  2  5/129 vs. 3/117  240 mg/day  3  1/125 vs. 2/120  300 mg/day 400 mg/day 480 mg/day  1 1 1  0/58 vs. 0/51 0/58 vs. 0/51 2/46 vs. 2/45  RR (95% CI) 0.54 (0.05, 5.69) Not estimable 0.25 (0.03, 2.22) 0.34 (0.01, 8.13) 1.55 (0.39, 6.23) 0.66 (0.11, 3.95) Not estimable Not estimable 0.98 (0.14, 6.65)  Five trials reported W D A E for verapamil (Chan 1997, Cushman 1998, Felictta 1992, Meeves 1994, Whelton 1992).  There was not enough power to achieve statistical  significance because of the limited amount of data.  214  3.11.3  Other c a l c i u m channel blockers  3.11.3.1  M i b e f r a d i l vs. placebo  T a b l e 99: Effect of mibefradil o n withdrawals due to adverse events Dose of Mibefradil  # of trials  6.25 mg/day  2  # of W D A E in treatment group vs. placebo group 2/92 vs. 0/87  12.5 mg/day  2  2/90 vs. 0/87  25 mg/day  3  1/128 vs. 0/129  50 mg/day  3  1/131 vs. 0/129  100 mg/day  3  2/132 vs. 0/129  150 mg/day  2  3/76 vs. 0/78  200 mg/day  1  4/37 vs. 0/36  RR (95% CI) 2.83 (0.30, 26.70) 2.89 (0.31, 27.29) 3.00 (0.13, 71.96) 3.00 (0.13,71.96) 2.82 (0.30, 26.61) 7.34 (0.39, 137.78) 8.76 (0.49, 157.12)  Three trials reported W D A E data for mibefradil (Bernink 1996, Bursztyn 1997, Oparil 1997). Though there was a trend of increasing W D A E with higher doses, there was not enough power to achieve statistical significance at each dose because of the limited amount of data.  Elowever, when all doses are pooled, there is a statistically significant  increase in W D A E compared with placebo (RR 3.90, 95% CI: 1.48, 10.28).  215  4. DISCUSSION This systematic review of the blood pressure lowering efficacy of CCBs used data from published randomized placebo-controlled trials of 3-12 weeks duration, in order to determine the dose-related effects on systolic and diastolic blood pressure, heart rate, and the number of withdrawals due to adverse events. One important aim was to determine i f a difference in the magnitude of B P lowering existed amongst the different subclasses of CCBs or amongst the drugs within each subclass. Another aim was to determine optimal dosing for each C C B drug so as to assist prescribing.  It is important to note that  conclusions about dose-response were limited to doses that were assessed in the studies included in this systematic review.  4.1  What methodological issues a n d potential sources of bias were encountered while conducting the systematic review?  Several methodological issues, from searching databases to extracting and analyzing data, surfaced during the process of conducting this systematic review. The search strategy utilized the O V I D search interface, which was generally easy to use. However, it was not possible to search suffixes such as "-dipine" which would have been helpful in identifying dihydropyridine-related articles. The search strategy was developed for high sensitivity at the expense of specificity and the process of sorting through irrelevant references was timeconsuming. The C E N T R A L (Cochrane Central Register of Controlled Trials) database is formed from specialized registers from Cochrane groups, together with references to clinical trials identified in M E D L I N E  and E M B A S E .  Thus, with its enhanced  specificity,  C E N T R A L may prove to be a time-saving alternative to searching all of the standard electronic databases.  216  This systematic review is based on data that are reported in publications and in some cases, provided by the authors through correspondence.  Inadequate reporting of trial  methodology and data was a frequent problem. Most crossover trials did not report precrossover data and thus, these trials were excluded. Because only 36% of trials reported S D of B P change, in the majority of trials the variances were imputed. Some trials only reported blood pressures from ambulatory blood pressure monitoring even though office blood pressures were measured as well. Attempts to obtain additional data by contacting authors were seldom successful. O f the authors who responded, more often than not the data were no longer available, especially i f over a decade had lapsed since the year of publication. Authors who worked for a pharmaceutical company tended to have easier access to clinical trial data. Some authors referred me to pharmaceutical and/or statistical analysis companies, but garnering additional data by this method was rarely successful. T o promote standardization of data reporting, many journals have endorsed the C O N S O R T statement, consisting of a checklist of items to be included when reporting a clinical trial (87). This checklist covers items that were commonly deficient in published R C T s included in this systematic review, such as method of randomization, flow of participants through each stage, and adverse events in each group. Errors in the reported data are a potential threat to the validity of a systematic review. Some trials reported standard deviations that were spuriously low and were more likely to be standard errors. The trials that reported standard deviations of B P change that were > 3 standard deviations away from the weighted mean value were discarded and imputed values were used. Clinical trial fraud is also a possible source of bias in the results, and although none of the publications of the included trials have ever been withdrawn, the  217  primary author of one included trial (Fiddes 1994) pleaded guilty to fraud for falsifying records and endangering patients while he was the head of a clinical research institute (377). Errors in the process of extracting data were minimized by having two independent reviewers extract data. Whenever possible, data were extracted from text and tables in the publications since extracting graphical data may introduce inaccuracies. Loss of blinding in a trial would bias the results in favor of CCBs over placebo. A n investigator's knowledge of the occurrence of common C C B side effects such as flushing and ankle edema in clinical trial participants may compromise blinding. However, none of the included trials assessed the success of blinding in patients or investigators. Overall, the percentage of patients completing each trial was in the range of 85100%. Because the drop-out rate in the included studies was low, attrition bias is unlikely. The systematic review was based on group averages from each trial, rather than analyzing  individual  patient  data.  The  INDANA  (Individual Data  Analysis  of  Antihypertensive intervention trials) project, which involves meta-analyses of individual patient data, has the potential advantages of increased data reliability and improved ability to analyze large amounts of data in subgroups, assess prognostic factors, and relate treatment effects to baseline characteristics (378). 4.1.1  P u b l i c a t i o n bias Another potential methodological problem was publication bias since only published  trials were assessed in this review. Funnel plot asymmetry can indicate publication bias but can also be due to clinical and methodological heterogeneity.  Funnel plot asymmetry was  observed for several C C B drugs individually, as well as dihydropyridines as a subclass. Thus, effect sizes calculated in this systematic review are likely to be overestimates. Ideally, clinical trial registries would minimize the problem of publication bias by allowing "negative" trials  218  that would otherwise be unpublished to be included in the review. The World Health Organization has formed the International Clinical Trials Registry Platform with the aims of ensuring that all clinical trials are registered and that a minimum set of results will be reported and made publicly available (379). There are several techniques that can be applied to adjust for publication bias in a meta-analysis (380), but currently there is no gold standard. parametric "trim-and-fill" method was applied.  In this review, the non-  For the dihydropyridine subclass, 7  "missing" studies in the SBP analysis and 5 "missing" studies in D B P analysis were "filled i n " to achieve a symmetrical funnel plot. The resulting effect sizes were not altered in a clinically or statistically significant degree compared with the original data; this method suggested the data had overestimated change in SBP by 0.4 m m H g and change in D B P by 0.3 m m H g . A post-hoc subgroup analysis was done to determine if the degree of reported blood pressure lowering differed in large trials (active treatment arms in the highest tertile) versus small trials (active treatment arms in the lowest tertile). dihydropyridines were analyzed separately.  Dihydropyridines and non-  For dihydropyridines, this analysis confirmed  that the effect sizes for change in SBP and D B P in the smaller trials were statistically significantly greater and more variable than those in the larger trials. The exaggerated effect size of the included trials with small sample size corroborates evidence of publication bias in addition to the asymmetrical funnel plots associated with several of the dihydropyridine drug analyses. Using this method one would accept the tertile of the largest trials as the closest to the treatment effect. This would lead to a reduction in the overall effect from -10.5 to -9.0 m m H g for systolic and from -6.9 to -5.4 m m H g for diastolic.  219  Either way there are likely to be small studies with littie or no reduction in blood pressure that have not been published and therefore are not included in this review. However, locating such unpublished studies is generally unfruitful and the inclusion of such studies may not reduce bias i f the data are provided by interested sources or i f their quality differs from that of published trials (381). For non-dihydropyridines, there was no statistically significant difference in effect size for SBP and D B P between small and large trials. Flowever, the trend was in the same direction and this analysis lacked power since there were far fewer non-dihydropyridine trials included in the review than dihydropyridine trials. Also, the non-dihydropyridine trials had larger sample sizes than the dihydropyridines on average (mean size of active treatment arm of 54 patients vs. 43 patients, respectively), although the difference was not statistically significant.  4.1.2.  Selection Bias  The method of patient recruitment is another potential source of bias, as studies could select for known responders to C C B s in previous trials to participate in new trials. Thus there are likely to be some patients who participated in several C C B trials. However, the degree of selection bias is difficult to quantify because the method of patient recruitment is usually not reported adequately.  In this systematic review, we searched for selection bias  by dividing up the trials into tertiles according to year of publication. The mean effect size of the oldest trials was compared with that of the most recent trials. If selection bias was present, we would expect the tertile with the most recent trials to have a greater effect size than that of the oldest trials.  When this was done there was no statistically significant  220  difference in effect size between the oldest and most recent trials. Therefore, we could not confirm any suggestion of evidence of selection bias in this systematic review.  4.2  What is the dose-related, b l o o d pressure lowering efficacy of each subclass? Is there a difference i n the best estimate of the magnitude of B P lowering effect of different subclasses of C C B s ? In this systematic review, 106 trials met the inclusion criteria and reported data on  13,878 patients (9513 receiving active treatment and 4365 receiving placebo), with a mean age of 55 years, mean baseline blood pressure of 158.2/101.6 m m H g and mean pulse pressure of 56.7 mm H g . Data were pooled for the dihydropyridine subclass and the non-dihydropyridine subclass by categorizing individual doses according to multiples of the manufacturer's starting dose (0.5x, l x , 2x, 4x). Both subclasses demonstrated a dose-response relationship, with an increasing magnitude in B P lowering as the dose increases from half the starting dose up to twice the starting dose. The maximal blood pressure lowering efficacy appeared to be attained at twice the recommended starting dose; at doses above this, there is no corresponding increase in blood pressure lowering response. For dihydropyridines, the best estimate of the maximal blood pressure lowering is -10/-7 m m H g , based on 2857 patients for SBP and 3076 patients for D B P . The best estimate of the maximal blood pressure lowering efficacy of the non-dihydropyridine subclass is -8/-6 m m H g , based on 1309 patients for SBP and 1632 patients for D B P . The pooled data have narrow confidence intervals, thereby allowing detection of even small differences between subclasses. The results of this systematic review reveal that dihydropyridines are statistically significantly more effective at lowering blood pressure than  221  non-dmydropyridines, although one could argue that this difference may not be clinically significant.  4.3  Is there a difference i n the best estimate of the magnitude of B P lowering effect of drugs i n each subclass?  Differences between C C B drugs within each subclass cannot be assessed accurately in this systematic review because the analysis was restricted to doses that were tested in the included clinical trials. For several drugs there were not enough data to define the doseresponse relationship. There was also high variability in the amount of available data for each dose of each drug. The effect sizes for drugs with littie data have very wide confidence intervals (Table 51).  4.4  What is the effect o n b l o o d pressure i n the placebo group i n short-term trials? The weighted mean changes in SBP and D B P in placebo groups across all included  trials with blood pressure data were -3.3 m m H g (SD 4.1; range -16.7 to 9) and -3.5 mm H g (SD 2.7; range -11.4 to 4.5), respectively. Because the placebo response was quite variable, it was important to subtract this effect for each trial in order to isolate the effect of the C C B drug. Hence, the inclusion criteria mandated that included trials have a parallel placebo arm. The placebo effect arises from the natural history of the condition, regression to the mean, and non-specific effects of treatment (382).  The variability associated with the  measurement of blood pressure itself may also contribute to the wide range of placebo responses observed in this systematic review.  222  4.5  Does the method of b l o o d pressure measurement affect the b l o o d pressure lowering efficacy of C C B s ?  Sources of error associated with blood pressure measurement could affect study entry and trial outcomes, thereby potentially imparting bias onto the results of a systematic review. Thus, sensitivity analyses were planned to ascertain i f the method of blood pressure measurement and the position of measurement affected drug efficacy. Seventy-two percent of the trials utilized sphygmomanometers and 6% of the trials used automatic devices to measure clinic blood pressures, while 22% of the trials did not report the instrument used. There were insufficient data to conduct a sensitivity analysis based on instrument used to measure B P . The effect of calcium channel blockers on ambulatory blood pressures as measured by 24-hour ambulatory blood pressure monitors was not assessed in this systematic review. Most of the blood pressure data extracted were sitting (48%), followed by supine (26%) and standing (16%) and average of supine/standing (4%) positions. The position of measurement was not reported for 6% of the trials with extractable B P data. There was no statistically significant difference in B P lowering efficacy among the different measurement positions.  4.6  Does trial quality affect the b l o o d pressure-lowering efficacy of c a l c i u m channel blockers?  Both the Jadad and Cochrane quality assessment scales were utilized in this review. One hundred (94.3%) of the included trials did not report allocation concealment, while the remaining six (5.7%) trials reported an adequate method of concealment.  Most included  trials (86 trials; 81.1%) were of moderate quality, while 19 (17.9%) were of low quality and  223  only one trial (0.9%) was of high quality. A sensitivity analysis removing low quality studies did not alter the effect sizes obtained in the C C B subclass analyses.  The relative  homogeneity of the quality of the studies is reflected in the screening criteria for inclusion into the systematic review. A l l included studies had to be randomized, double-blinded and placebo-controlled. Because the Jadad scoring criteria match these screening criteria, this scale was not particularly useful in this review. The Cochrane method of assessing trial quality with respect to allocation concealment was also not useful because most studies did not report these details.  It is worth noting that the quality of a trial's reporting of the  methodology and the quality of the trial results are not always in accord. The accuracy of blood pressure measurement is the most important factor affecting the quality of the included studies. However, this factor is not taken into account in the Jadad and Cochrane quality assessment scales.  4.7.  Was there a difference i n b l o o d pressure lowering efficacy at trough vs. peak? Eighty-four trials reported trough blood pressures and 3 trials reported peak blood  pressures, while 19 trials did not report timing of blood pressure measurement.  The small  number of trials with peak BPs precludes formal comparison of effect sizes at trough versus peak timing. A future systematic review of trials with 24-hour blood pressure monitoring would best address this question because this measurement technique allows both trough and peak data to be obtained for each patient.  224  4.8  D i d funding source affect the reported b l o o d pressure lowering efficacy of CCBs? Relationships between funding source and trial outcome in R C T s have been assessed  in several studies.  A published systematic review comparing the outcomes of industry-  sponsored vs. nonindustry-sponsored original research showed a statistically significant association between industry-sponsorship of RCTs and pro-industry results [OR 4.14 (95% C L 2.73-6.32)] (383). The majority of trials included in this systematic review (60; 57%) did not report the source of funding. O f the trials that did report funding source, 45 (98%) had an industry funding source, with 40 trials (87%) with potential bias in favor of the C C B drug being tested, and 5 trials (11%) with potential bias against.  Only one trial (2%) was funded by a  government agency. Sensitivity analysis removing the trials with potential bias against the C C B drug and government-funded trials did not change the results. This analysis is not adequately powered because of the general lack of reporting.  4.9  Does age affect the b l o o d pressure lowering efficacy of C C B s ? Eight studies had inclusion criteria for older patients, defined as cutoff age > 60  years (Chan 1997, Rizzini 1991, BarbagaUo 2000, Ninci 1997, Fogari 1999, Bursztyn 1997, Scuteri 1992, Paolisso 1991). A n additional four trials had different age criteria but still had a mean age of > 60 years (Fiddes 1994, Black 2001, van Ree 1996, Gerritsen 1998). T w o trials reported data separately for older patients and younger patients (Fagan 1997, Fiddes 1994). However, with such little data a subgroup analysis of the effect of CCBs on older vs.  225  younger patients was not feasible. A systematic review of individual patient data would be more suitable to assess this question, as more data would be available in such an analysis.  4.10  Does co-morbidity alter the b l o o d pressure lowering efficacy of C C B s ? A subgroup analysis of hypertensive patients with co-morbidity was not feasible in  this systematic review because of the lack of such patients in the included studies.  Two  nitrendipine studies had inclusion criteria of Type 2 diabetes mellitus. One nicardipine study was performed in hospitalized patients.  One trial assessed lidoflazine in post-infarction  patients. Otherwise, all other trials did not select for patients with co-morbidity. Essentially, most trials excluded patients with major cardiac, hematologic, renal, hepatic, or endocrine disease.  4.11  Does b l o o d pressure lowering efficacy of C C B s differ for isolated systolic hypertension vs. diastolic or systo-diastolic hypertension? Calcium channel blockers have been used commonly a first-line therapy for isolated  systolic hypertension, a guideline that was primarily based on the results of the "Systolic Hypertension in Europe" (SYST-Eur) trial.  In the SYST-Eur study, nitrendipine (with  optional add-on enalapril and/or hydrochlorothiazide) in comparison with placebo showed a reduction in the risk of stroke (RR 0.61, 95% CI 0.43-0.87) and cardiovascular events (RR 0.71, 95% CI 0.57-0.87), but not in total mortality and C A D (70). This systematic review included 3 trials with inclusion criteria of isolated systolic hypertension (Black 2001, Barbagallo 2000, Paolisso 1991). With such littie data, a subgroup analysis comparing the effect sizes between trials with inclusion criteria of I S H vs. diastolic hypertension was not feasible.  226  4.12  H o w do the direct comparisons between doses differ from the indirect comparisons? Direct comparisons were based on much less data than the indirect comparisons.  There was general agreement between the direct and indirect comparisons, but the confidence intervals were fairly wide.  227  4.13  F o r each C C B d r u g , do the manufacturer's starting doses coincide w i t h the lowest effective dose as determined by this systematic review?  T a b l e 100: C o m p a r i s o n of manufacturer's recommended starting doses and lowest effective doses determined i n this systematic review Lowest effective dose Manufacturer's Drug recommended starting ( m g / d a y ) dose ( m g / d a y ) 5mg 2.5 mg Amlodipine Barnidipine  10 mg  30 mg  Darodipine  n/a  100 mg  Diltiazem  120-240 mg  120 mg  Felodipine  5mg  2.5 mg  Isradipine  5 mg  1 mg  Lacidipine  2-4mg  4mg  Lercanidipine  10 mg  10 mg  Lidoflazine  n/a  n/a  Manidipine  10 mg  10 mg  Nicardipine  60 mg  40 mg  Nifedipine  20-30 mg  20 mg  Nilvadipine  8mg  8mg  Nisoldipine  10 mg  10 mg  Nitrendipine  5-20 mg  10 mg  Pranidipine  n/a  2mg  Tiapamil  n/a  N o doses showed statistically significant difference from placebo 180 mg  Verapamil  180-240 mg  228  There is general agreement between manufacturer's recommended dose and the lowest effective dose determined by this systematic review. However, the lowest effective doses  for  amlodipine,  felodipine,  isradipine  and  nicardipine  were lower  than  the  manufacturer's recommended starting doses.  4.14  What is the effect of C C B s on B P variability?  To determine the effect  of C C B treatment on B P variability, the  endpoint  variabilities of the C C B group were compared with the placebo group. There were a lot of data to assess this parameter and there was no statistically significant difference between C C B and placebo for SBP or D B P , demonstrating with a high degree of certainty that CCBs do not change B P variability. In absolute terms, the variability of SBP is statistically significantly different than that of D B P . However, in terms of coefficient of variation, which is a measure of the variability relative to the mean value, there is no statistically significant difference in the variabilities of SBP and D B P . Thus, it can be concluded that SBP and D B P vary to the same degree. The variability at baseline appears to be affected by blood pressure criteria for entry into the trials. SBP at baseline is statistically significandy lower in the 3 trials that used SBP as entry criteria compared to the 72 trials using D B P - or mixed S B P / D B P criteria.  The  baseline variabilities in D B P were similar across all entry criteria. However, in the trials with D B P entry criteria, the baseline S D values were statistically significandy lower than the endpoint values in both the C C B treatment and placebo groups. This demonstrates that the baseline D B P variabilities are likely to be spuriously low in trials with D B P entry criteria. Entry criteria may falsely lower the magnitude of baseline variability because it is a truncated  229  rather than a normal distribution. This effect would be further magnified i f many patients with BPs near the cut-off value are enrolled into trials. The blood pressure variability of a trial gives an indication of the reliability of the data since expected values can be compared with the reported values. Based on the baseline variabilities of treatment and placebo groups, the best estimate of the variability in SBP from trials with systo-diastolic hypertension is 14.4 mm H g (SD 3.2). However, the best estimate of the variability in D B P cannot be determined from baseline measures in this review because the majority of the trials had D B P entry criteria.  The best estimate for D B P  variability is based on end of treatment values, 8.2 m m H g (SD 2.4). This review assesses variability based on mean values from treatment groups, which accounts for both inter- and intra-individual variability. Intra-individual B P variability can be assessed independendy in cross-over studies or from data from 24-hour blood pressure monitoring. The variability of the change in B P in both SBP and D B P is not statistically significandy different in the C C B treatment group compared with the control group, ft can be concluded that the blood pressure responses to C C B treatment and placebo treatment are both highly variable (SD of S B P / D B P change of 13.5/7.8 m m H g and 14.1/7.9 mm H g , respectively), and the C C B effect to lower B P has no effect on that variability. Many trials did not report the S D of the change in B P , necessitating imputing these values.  However, the mean reductions in blood pressure were insensitive to the strategy  used to impute missing variances — that is, whether the hierarchy for imputation was used or the weighted mean S D of B P change from all trials reporting this parameter was used.  230  4.15  What is the effect of C C B s o n pulse pressure?  Pulse pressure is an independent risk factor for cardiovascular disease, although it was not reported as a primary or secondary outcome in any of the included trials. The weighted mean change in pulse pressure was calculated for dihydropyridines and nondihydropyridines at twice the starting dose and above, from trials reporting trough SBP and D B P data.  These parameters were compared with that of placebo, which was calculated  from all trials reporting both SBP and D B P , regardless of timing of B P measurement.  The  results demonstrate that placebo does not affect pulse pressure (weighted mean 0.3 mm H g , 95% CI: -0.2, 0.8), while a small reduction in pulse pressure is observed with both dihydropyridines (-3.4 m m H g ; 95% CT. -4.3, -2.5) and non-dihydropyridines (-2.4 mmHg; 95% CI: -3.7, -1.1). There is no statistically significant difference between dihydropyridines and non-dihydropyridines in change in pulse pressure (p = 0.2). In a sensitivity analysis, studies with isolated systolic hypertension as entry criteria were removed from the analysis. There were only two such studies (Barbagallo 2000 and Paolisso 1991), both of which had been included in the pulse pressure calculation for placebo only, because the timing of blood pressure measurement was not reported. Without the I S H studies, the effect size for the placebo group remains similar (0.4 mm H g ; 95% CI: -0.1, 0.9). Placebo does not alter pulse pressure most likely because none of the factors that contribute to the placebo response has any effect on blood pressure or its regulation. This is evidence against the placebo response being due to a psychological effect on blood pressure.  231  4.16  Is there any evidence of a dose-response relationship w i t h respect to change i n heart rate? Acutely,  short-acting  CCBs  are  associated  with  increased  heart  rate  and  norepinephrine levels. With chronic dosing, the effect of C C B s on heart rate appears to be small and clinically insignificant, based on the 47 (44%) trials reporting heart rate data in this review. However, this result may be due to bias from lack of reporting, since many trials did not report this outcome. For dihydropyridines, doses greater than or equal to 1.5 times the recommended starting dose are associated with a statistically significant increase in heart rate of 1.4 beats per minute (95% CI: 0.60, 2.2) compared with placebo. For non-dihydropyridines, there are insufficient data to obtain a valid estimate of the effect on heart rate. However, the limited data from this systematic review demonstrated that non-dihydropyridines at doses 1-3 times the starting dose were associated with a statistically significant decrease of 2.6 beats per minute (95% Cf: -3.85, -1.36) compared with placebo.  4.17  Is there any evidence of a dose-reponse relationship w i t h respect to withdrawals due to adverse events?  The number of withdrawals due to adverse events in each dosage group was reported in 64 (60%) of the included trials. However, the type and severity of adverse events were not consistendy reported. Reports of adverse effects from published and unpublished sources other than randomized controlled trials may provide better information on the long term safety of C C B drugs. Ffowever, some rudimentary conclusions can be gleaned from the short-term RCTs included in this systematic review.  For each drug considered  independendy, there were not enough data to make an accurate estimate of W D A E .  232  However, when grouped as a subclass, the cLihydropyridines are associated with a statistically significant increase in W D A E compared with placebo. There also appears to be a doseresponse relationship, with increasing relative risk values with increasing dose, starting with a non-significant difference at the starting dose, a R R of 1.76 (95% CI 1.21, 2.58) at twice the starting dose, and a R R of 3.91 (95% CI 2.19, 6.99) at >3 times the starting dose. The nondihydropyridine trials did not show a statistically significant difference in W D A E compared with placebo but there is insufficient evidence to make a valid estimate of their effect on WDAE.  4.18  Can the magnitude of blood pressure lowering efficacy of calcium channel blockers be linked to their mechanism of action?  Hypertension is associated with elevated concentration of intracellular calcium ions, which can arise from increased transport from extracellular calcium stores, from increased release from intracellular stores, or both. The exact mechanism of action of calcium channel blockers is not known but they probably reduce blood pressure by vasodilatory and cardiodepressant effects resulting from a reduction of calcium entry through the L-type calcium channels on the arterial vasculature and cardiac tissue. The blood pressure lowering efficacy of CCBs is quite modest in relation to the high degree of variability associated with blood pressure measurement.  Human physiology is  designed to maintain blood pressure levels through several homeostatic mechanisms. For instance, baroreceptors in carotid sinuses and aortic arch detect changes in arterial pressure, and they buffer such changes by activating reflex responses in the heart, blood vessels and kidney (384). Chronic hypertension involves resetting of the baroreceptor reflex to higher pressures.  233  A t the cellular level, there are many players involved in the regulation of vascular tone other than L-type calcium channels, such as several types of potassium channels, chloride channels, store-operated calcium channels, and stretch-activated cation channels (385). L-type calcium channels are regulated by the membrane potential, which in turn is determined essentially by potassium channels.  Stretch-actived cation channels provide a  route of entry for calcium into vascular muscle cells, even when dihydropyridine-sensitive channels are blocked (385). Store-operated channels allow calcium entry when intracellular stores are low. In this review a small statistically significant difference in the BP-lowering efficacy between dihydropyridines and non-dihydropyridines was demonstrated. However, in view of the complexity of blood pressure regulation and the action of CCBs it is not possible to rationally speculate as to why this may be.  4.19  H o w c a n the b l o o d pressure lowering efficacy of c a l c i u m channel blockers i n short-term trials be related to their effects o n mortality a n d morbidity outcomes i n long-term trials?  There is current debate regarding the effect of CCBs on morbidity and mortality outcomes in long-term trials.  Pahor et al.'s recent systematic review of randomised  controlled trials found no statistically significant difference in the reduction of SBP or D B P between calcium antagonists and other drugs (diuretics, beta blockers, A C E inhibitors, clonidine) (55). However, the review also demonstrated that C C B s were associated with a significandy higher risk of acute myocardial infarction (odds ratio [OR], 1.2; 95% CI 1.111.43), congestive heart failure (OR 1.25; 95% CI 1.07-1.46) and combined major cardiovascular events (OR 1.10; 95% CI 1.02-1.18) compared to the other drug classes. This  234  suggests that that clinical outcomes associated with C C B treatment are not dependent solely only on the magnitude of blood pressure reduction, but also on other factors. This claim has been refuted but more studies need to be done to clarify this area of controversy. Recent R C T s have demonstrated increased risk of heart failure with C C B treatment (59, 60). Another systematic review is being conducted currendy to assess adverse cardiac effects associated C C B therapy (386). The blood pressure lowering efficacy observed in short-term trials of C C B monotherapy cannot be compared direcdy to long-term trials in which other drugs are allowed to be added to C C B treatment in order to meet blood pressure targets. Nonetheless, it is worth noting the magnitude of BP-lowering in long term, placebo-controlled C C B trials, where other drugs could be added i f blood pressure targets were not reached.  In the  SYST-Eur trial (70), first-line nitrendipine lowered B P by -10/-4.5 mm H g (placebocorrected), and a similar magnitude of blood pressure lowering was found with first-line nitrendipine in the SYST-China study (-9/-3 mm Hg) (69). The choice between different classes of antihypertensive drugs should be based on morbidity and mortality data.  Currendy, the first-line pharmacological treatment of  hypertension is low-dose thiazide diuretics, which have been shown to reduce cardiovascular events, cardiovascular mortality and total mortality compared with placebo. Thiazdes have demonstrated a mean blood pressure reduction of 10/4 mm H g in a previous systematic review (81), a magnitude that is similar to the calcium channel blocker class. However, thiazide diuretics confer a reduction in heart failure outcomes as compared with CCBs. One can speculate that the ability of thiazides to reduce pulse pressure more than CCBs (6 mm H g for thiazides vs 3 m m H g for dihydropyridines and 2 mm H g for non-dihydropyridines) is a possible explanation for this difference.  Further investigation of pulse pressure  235  differences in long term head-to-head trials between antihypertensive classes could be done to test this hypothesis. Based on blood pressure lowering efficacy alone, the dihydropyridines appear to be better  than  non-dihydropyridines.  However,  this  review  does  not  suggest  that  dihydropyridines are better at lowering blood presssure than thiazides and considering the data on morbidity and mortality outcomes, calcium channel blockers should remain as second- or tlurd-Iine agents for management of primary hypertension.  236  5.  CLINICAL IMPLICATIONS  The findings of this systematic review represent the best available evidence about the blood pressure lowering efficacy of calcium channel blockers in adults with primary hypertension. Several important conclusions are relevant to the clinical use of this class of drugs.  1. Dihydropyridines reduce blood pressure to a greater degree than non-dihydropyridines. The  best  estimate  of the  maximal  blood  pressure  lowering  efficacy  dihydropyridine subclass is -10 m m H g for SBP and -7 m m Pig for D B P .  of  the  The best  estimate of the maximal blood pressure lowering efficacy of the non-dihydropyridine subclass is -8 m m Pfg for SBP and -6 mm Pfg for D B P .  2. Publication bias was present in this systematic review, leading to an overestimate of the effect size; the magnitude of this overestimate is likely between 0.4 m m H g and 1.5 m m H g for SBP 0.3 and f .5 m m H g for D B P .  3. There was a dose-response relationship for the blood pressure lowering efficacy of CCBs.  For dihydropyridines and non-dihydropyridines, half the starting dose lowered  blood pressure less and twice the starting doses lowered blood pressure more than starting doses.  Maximal blood pressure lowering appears to be achieved at twice the  manufacturer's recommended starting dose. Further dose escalation does not enhance blood pressure lowering efficacy despite manufacturers' suggestions otherwise.  237  4. Regarding blood pressure variability: a)  Treatment of primary hypertension with calcium channel blockers does not affect blood pressure variability compared with placebo.  b) Systolic and diastolic blood pressures vary to the same degree when the variability is expressed as coefficient of variation. c)  Blood pressure criteria for entry into randomized controlled trials results in a reduction in the estimate of the baseline variability of the corresponding parameter.  5. Regarding pulse pressure: A)  Placebo response does not affect change in pulse pressure.  B)  Both dihydropyridines and non-dihydropyridine CCBs reduce pulse pressure by about 3 m m H g .  6.  Calcium channel blockers given for 3-12 weeks have a small and probably clinically insignificant effect on heart rate. a)  For dihydropyridines, doses greater than or equal to 1.5 times the  manfacturer's  recommended starting dose are associated with a statistically significant increase in heart rate of 1.4 beats per minute. b) For non-dihydropyridines, doses 1-3 times the manufacturer's recommended starting dose are associated with a statistically significant decrease of 2.6 beats per minute.  7. A l l doses combined, dihydropyridines increase withdrawals due to adverse events compared with placebo at starting doses (RR 1.7 [95% CI: 1.3, 2.1]; Absolute risk increase 2%; number need to harm = 50). A dose-response relationship was present,  238  with a R R of 1.76 (95% C I 1.21, 2.58) at 1.5-3 times the starting dose, and a R R of 3.91 (95% C I 2.19, 6.99) at >3 times the starting dose.  The non-dihydropyridine trials did  not show a statistically significant difference in W D A E compared with placebo but this was probably due to lack of reporting.  Dose titration of calcium channel blockers above the recommended starting dose is associated with a minimal benefit in terms of increase in blood pressure lowering efficacy, and increased harm, in terms of increased withdrawals due to adverse effects and greater cost.  239  6. I M P L I C A T I O N S F O R F U T U R E R E S E A R C H  1.  Because publication bias was found in this review, the published data on the short-term blood pressure lowering efficacy of calcium channel blockers represents an incomplete data set. Thus, all clinical trials should be registered and reported in full.  2. There is a need to improve the quality of reporting of details in publications of clinical trials measuring blood pressure lowering efficacy. A s systematic reviews are secondary research, the validity of the results depends on the quality of its component trials. A)  Baseline and endpoint blood pressures, the mean change  from  baseline and the standard deviations of all of these parameters must be reported for all randomized groups and provided in a table. B)  The  number  of  serious  adverse  events,  deaths,  number  of  withdrawals due to adverse events in each group, reasons  for  withdrawal, and the time point at which withdrawals occurred for each group must be reported in all trials. C)  Complete information must be given regarding % of patients completing the trial as well as the number of dropouts and reasons thereof.  Methodological details regarding randomization, allocation  concealment and blinding should be reported adequately to facilitate assessment of trial quality. Complete baseline characteristic details are also necessary to evaluate the effect size within trials and to validate the combining of data across different trials in meta-analyses.  240  3.  Improving the search capabilities and indexing systems of medical literature databases and clinical trial registries will facilitate more complete and efficient searching for future systematic reviews and updates.  The Cochrane Central Register of Controlled Trial  ( C E N T R A L ) is a highly sensitive, specific and up-to-date database which provides a time-saving alternative to searching all of the standard electronic databases.  4.  More clinical trials exploring a wider range of doses of calcium channel blockers need to be conducted.  5. A systematic review of clinical trials with head-to-head comparisons of different doses of each calcium channel blocker drug should be performed in order to validate the relative blood pressure-lowering efficacies of the various drugs within this class.  6.  A systematic review of cross-over trials of CCBs in hypertension should be performed to validate the blood pressure-lowering efficacy of CCBs as determined by this systematic review.  Asssessment of end-of-treatment blood pressure standard deviations from  cross-over studies would give an estimate of the within-patient blood pressure variability.  7. A systematic review of the adverse effects of calcium channel blockers should be performed in order to explore further the nature and severity of adverse events leading to withdrawals from clinical trials.  8.  More long-term trials of head-to-head comparisons of calcium channel blocker drugs with other drug classes using doses with comparable blood pressure control need to be  241  performed in order to determine whether there are class-specific effects on morbidity and mortality.  242  7. R E F E R E N C E S 1.  Beevers D G , MacGregor G A . Hypertension in Practice, 3 edition. London: Martin D u n k z Ltd; 1999.  2.  Joffres M R , Hamet P, Rabkin SW, Gelskey D , Hogan K , Fodor G for the Canadian Heart Health Surveys Reseach Group. Prevalence, control and awareness of high blood pressure among Canadian Adults. C M A J 1992; 146: 1997-2005.  3.  World Flealth Organization, International Society of Flypertension Writing Group. 2003 World Flealth Organization/International Society of Hypertension (ISH) statement on management of hypertension. J Flypertension 2003; 21:1983-1992.  4.  Seedat Y K . The limits of antihypertensive therapy — lessons from Third World to First. Cardiovascular Journal of South Africa 2001; 12(2): 94-100.  5.  National High Blood Pressure Education Program Coordinating Committee. The sixth report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. A r c h Intern M e d 1997; 157:2413-2446.  6.  Rose G . Epidemiology. In: Marshal A J and Barrett D W , editors. The Flypertensive Patient. Kent, England: Pitman Medical; 1980. p. 1-21.  7.  Kaplan N M . Hypertension: prevalence, risks, and effect of therapy. A n n Intern M e d 1983; 98: 705-709.  8.  Port S, Demer L , Jennrich R, Walter D , Gardinkel A . Systolic blood pressure and mortality. Lancet 2000; 355: 175-180.  9.  Prospective Studies Collaboration. Age-specific relevance of usual blood pressure to vascular mortality: a meta-analysis of individual data for one million adults in 61 prospective studies. Lancet 2002; 360: 1903-1913.  10.  Feldman R D , Campbell N , Larochelle P, Bolli P, Burgess E D , Carruthers S G for the Task Force for the Development of the 1999 Canadian recommendations for the management of hypertension. 1999 Canadian recommendations for the management of hypertension. C M A J 1999; 161 (12 Suppl): S1-S22.  11.  Parati G , . Mancia G . Blood pressure variability as a risk factor. Monitoring 2001; 6:341-347.  12.  Sander D , Kukla C, IClingelhofer J, Winbeck K , Conrad B . Relationship between circadian blood pressure pattersn and progression of early carotid atherosclerosis: a 3year follow-up study. Circulation 2000; 102:1536-1541.  13.  Mitchell G F , Moye L A , Braunwald E , Rouleau J-L, Bernstein et al. Sphygmomanometricalfy determined pulse pressure is a powerful independent  r  Blood Pressure  243  predictor of recurrent events after myocardial infarction in patients with impaired left ventricular function. Circulation 1997; 96:4254-4260. 14.  Verdecchia P, Schillaci, G , Borgioni C, Ciucci A , Pede S, Porcellati C. Ambulatory pulse pressure: a potent predictor of total cardiovascular risk in hypertension. Hypertension 1998; 32(6):983-988.  15.  Glynn RJ, Chae C U , Guralnik J M , Taylor J O , Hennekens C H . Pulse pressure and mortality in older people. A r c h Intern M e d 2000; 160: 2765-2772.  16.  Gillman M W , Kannel W B , Belanger A , D'Agostino R B . Influence of heart rate on mortality among persons with hypertension: The Frarningham Study. A m Heart J 1993; 125:1148-1154.  17.  Chobanian A V , Bakris G L , Black H R et al and the National High Blood Pressure Education Program Coordinating Committee. The seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. J A M A 2003; 289(19): 2560-2572.  18.  O'Rourke M . Arterial stiffness, systolic blood pressure, and logical treatment of arterial hypertension. Hypertension 1990; 15: 339-347.  19.  Baumbach G L . Chapter A 6 3 - Mechanisms of vascular remodeling, In Izzo J L , Jr and Black H R , editors. Hypertension Primer, 3 ed. Dallas: American Heart Association; 2003. p. 180-183. rd  20.  Mulvany M J . Effects of angiotensin converting enzyme inhibition on vascular remodelling of resistance vessels in hypertensive patients. Journal of Hypertension Suppl. 1996; 14(6): S21-24.  21.  Lyons D . Impairment and restoration of nitric oxide-dependent vasodilation in cardiovascular disease. International Journal of Cardiology 1997; 62 (Suppl 2): S101S109.  22.  Waeber B , Brunner H R . The multifactorial nature of hypertension: the greatest challenge for its treatment? Journal of Flypertension 2001; 19(suppl. 3) : S9-S16.  23.  Arnett D K , Boerwinkle E , Davis B R , Eckfeldt J, Ford C E , Black H . Pharmacogenetic approaches to hypertension therapy: design and rationale for the Genetics of Flypertension Associated Treatment (GenHAT) study. The Pharmacogenomics Journal 2002; 2: 309-317.  24.  Jiirgens G , Graudal N A . Effects of low sodium diet versus high sodium diet on blood pressure, renin, aldosterone, catecholamines, cholesterols, and triglyceride. The Cochrane Database of Systematic Reviews 2006 Issue 2.  244  25.  Hooper L , Barrier! C, Davey Smith G , Ebrahim S. Advice to reduce dietary salt for prevention of cardiovascular disease. The Cochrane Database of Systematic Reviews 2006 issue 2.  26.  Thompson P D , Buchner D , Pina IL, Balady G J , Williams M A et al. Evercise and physical activity in the prevention and treatment of atherosclerotic cardiovascular disease. Circulation 2003; 107: 3109-3116.  27.  Moser M . Historical perspective on the management of hypertension. The American Journal of Medicine. 1986; 80(Suppl. 5B): 1-11.  28.  McAlister F A , Campbell N R C , Zarkne K , Levine M , Graham I D . The management of hypertension in Canada: a review of current guidelines, their shortcomings and implications for the future. C M A J 2001; 164(4): 517-522.  29.  Wright J M , Lee C H , Chambers G K . Systematic review of antihypertensive therapies: does the evidence assist in choosing a first-Une drug? C M A J 1999; 161(l):25-32.  30.  Psaty B M , Lumley T, Furberg C D , Schellenbaum G , Pahor M , Alderman M H , Weiss NS. Health outcomes associated with various antihypertensive therapies used as firstline agents: a network meta-analysis. J A M A 2003; 289(19):2534-44.  31.  Naqvi N H , Balufox M D . Blood Pressure Measurement: A n Illustrated History. Carnforth, Lanes, U K : Parthenon Pubkshing; 1998.  32.  Nara A R , Burns M P , Downs W G . SpaceLabs Inc.; 1989. p. 41-93.  33.  McAlister F A , Straus S E . Chapter 2: What is this person's blood pressure? In Mulrow C D , editor. Evidence-based Ffypertension. London: B M J Books; 2001. p. 9-27.  34.  O'Brien E T , O'Malley K . Blood Pressure Measurement. New York: Elsevier Science; 1991.  35.  Gerin W , Marion R M , Friedman R. et al. H o w should we measure blood pressure in the doctor's office? Blood Pressure Monitoring 2001; 6:257-262.  36.  Myers M G . Use of trough:peak ratio in the assessment of antihypertensive drug therapy. Canadian Journal of Cardiology 1994; 10 (Suppl. D): 17D-20D.  37.  Rang H P , Dale M M , Ritter J M . Livingstone; 1999.  38.  Berridge M . Elementary and global aspects of calcium signalling. J. Physiol 1997; 499: 291-306  39.  Katz A M . Calcium channel diversity in the cardiovascular system. J A C C 1996; 28(2): 522-529.  Blood Pressure. Redmond, Washington:  Pharmacology. 4  th  ed. Edinburgh: Church  245  40.  Oates J A , Brown N J . Chapter 33 - Antihypertensive agents and the drug therapy of hypertension, In Hardman J G , Limbird L E , Gilman A G , editors. Goodman and Gilman's The Pharmacological Basis of Therapeutics. 10 ed. London: McGraw Hill; 2001. p. 871-900. th  41.  Vanhoutte P M . The Expert Committee of the World Plealth Organization on Classification of Calcium Antagonists: the viewpoint of the raporteur. Am J Cardiology 1987; 59:3A-8A.  42.  Triggle D J . Sites, mechanisms of action, and differentiation of calcium channel antagonists. A m J Hypertension 1991; 4(7) Part 2: 422S-429S.  43.  Hockerman GPI et al. Molecular determinants of drug binding and action on L-type calcium channels. A n n u Rev. Pharmacol Toxicol 1997; 37: 361-396.  44.  Canadian Pharmacists Association. Compendium of Pharmaceuticals and Specialties. [Online]. 2006 [cited 2006 September 14] Available from: U R L : http://ecps.pharmacists.ca/CPPIA/main.htm  45.  Frishman WPI. Mibefradil: a new selective T-channel calcium antagonist for hypertension and angina pectoris. J Cardiovasc Pharmacol Therapeut 1997; 2(4):321330.  46.  Roche Laboratories announces withdrawal of Posicor from the market. F D A Talk Paper [Online] 1998 June 8 [cited 2006 July 21]. Available from: U R L : http://wAvw.fda.gov/bbs/topics/ANSWERS/ANS00876.html  47.  Borchard U . Calcium antagonists in comparison: view of the pharmacologist. Journal of Cardiovascular Pharmacology 1994; 24(Suppl.2):S85-S91.  48.  IMS Global Insights - Lipitor Leads the Way in 2003. [Online]. 2004 [cited 2006 July 21]; Available from: URL:http://www.imshealtii.com/web/content/0.3148.64576068 63872702 7026099 8 70960214.00.html  49.  Psaty B M , Heckbert SR, Koepsell T D et al. The risk of myocardial infarction associated with antihypertensive drug therapies. J A M A 1995; 274(8):620-625.  50.  Pahor M , Guralnik J M , Corti M C , Foley D J , Carbomn P, Plavlik RJ. Long-term survival and use of antihypertensive medications in older persons. J A m Geriatr Soc 1995; 43(11):1191-1197.  51.  Furberg C D , Psaty B M , Meyer J V . Nifedipine: dose-related increase in mortality in patients with coronary heart disease. Circulation 1995; 92:1326-1331.  246  52.  Alderman M H , Cohen H , Roque R, Madhavan S. Effect of long-acting and shortacting calcium antagonists on cardiovascular outcomes in hypertensive patients. Lancet 1997; 349: 549-598.  53.  Kizer JR, Kimmel SE. Epidemiologic review of the calcium channel blocker drugs: an up-to-date perspective on the proposed hazards. A r c h Intern Med. 2001; 161:11451158.  54.  Stelfox H T . Chua G . O'Rourke K . Detsky A S . Conflict of interest in the debate over calcium-channel antagonists. N Engl J M e d 1998; 338(2):101-6.  55.  Pahor M , Psaty B M , Alderman MPI, Applegate W B , Williamson J D , Cavazzini C, Furberg C D . Health outcomes associated with calcium antagonists compared with other first-line antihypertensive therapies: a meta-analysis of randomized controlled trials. Lancet 2000; 356: 1949-1954.  56.  Dahof B , Sever PS, Poulter N R , et al. for the A S C O T investigators. Prevention of cardiovascular events with an antihypertensive regimen of amlodipine adding perindopril as required versus atenolol adding bendroflumethiazide as required, in the Anglo-Scandinavian Cardiac Outcomes Trial-Blood Pressure Lowering A r m ( A S C O T B P L A ) : a multicentre randomised controlled trial. Lancet 2005; 366:9489: 895-906.  57.  L i u L , Zhang Y , L i u G . for the F E V E R Study Group. The Felodipine Event Reduction ( F E V E R ) Study: a randomized long-term placebo-controlled trial in Chinese hypertensive patients. J Hypertens 2005; 23:2157-2172.  58.  Estacio R O , Jeffers B W , Hiatt W R , Biggerstaff SL, Gifford N , Schner R W . The effect of nisoldipine as compared with enalapril on cardiovascular outcomes in patients with non-insulin-dependent diabetes and hypertension. N Engl J M e d 1998; 338(10):64552.  59.  The A L L F I A T Officers and Coordinators for the A L L H A T Collaborative Research Group. Major outcomes in high-risk hypertensive patients randomized to angiotensin-converting enzyme inhibitor or calcium channel blocker vs diuretic: The Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial ( A L L H A T ) . J A M A 2002; 288:2981-2997.  60.  Black FIR, Elliot W F , Grandits G . et al for the C O N V f N C E Research Group. Principal results of the Controlled Onset Verapamil Investigation of Cardiovascular E n d Points ( C O N V I N C E ) trial. J A M A 2003; 289(16): 2073-2082.  61.  Tatti P. Pahor M . Byington RP. D i Mauro P. Guarisco R . Strollo G . Strollo F. Outcome results of the Fosinopril Versus Amlodipine Cardiovascular Events Randomized Trial ( F A C E T ) in patients with hypertension and N I D D M . Diabetes Care 1998. 21 (4):597-603.  62.  Hansson L , Zanchetti A , Carruthers S G , Dahlof B , Elmfeldt D , Julius S, Menard f,Rahn KFI, Wedel FI, Westerling S. Effects of intensive blood-pressure lowering and  247  low-dose aspirin in patients with hypertension: principal results of the Hypertension Optimal Treatment ( H O T ) randomized trial. H O T Study Group. Lancet 1998; 351(9118):1755-62. " 63.  Brown MJ, Palmer C R , Castaigne A , de Leeuw P W , Mancia G , Rosenthal T, Ruilope L M . Morbidity and mortality in patients randomized to double-blind treatment with a long-acting calcium-channel blocker or diuretic in the International Nifedipine GITS study: Intervention as a Goal in Hypertension Treatment (INSIGHT). Lancet 2000; 356(9227):366-72.  64.  Pepine CJ, Handberg E M , Cooper-DeHoff R M , et al. for the I N V E S T Investigators. A calcium antagonist vs a non—calcium antagonist hypertension treatment strategy for patients With coronary artery disease: The International Verapamil-Trandolapril Study (INVEST): A Randomized Controlled Trial. J A M A 2003; 290 (21): 2805-2816.  65.  Borhani N O , Mercuri M , Borhani P A , Buckalew V , Canossa-Terris M , Carr A A , et al. Final outcome results of the Multicenter Isradipine Diuretic Atherosclerosis Study (MIDAS). A randomized controlled trial. J A M A 1996; 276(10):785-91.  66.  National Intervention Cooperative Study in Elderly Hypertensives Study Group. Randomized double-blind comparison of a calcium antagonist and a diuretic in elderly hypertensives. Hypertension 1999; 34(5):1129-33.  67.  Flansson L , Fledner T, Lund-Johansen P, Kjeldsen S E , Lindholm L H , Syvertsen J O et al. Randomized trial of effects of calcium antagonists compared with diuretics and beta-blockers on cardiovascular morbidity and mortality in hypertension: the Nordic Diltiazem ( N O R D I L ) study. Lancet 2000; 356(9227):359-65.  68.  Hansson L , Lindholm LFI, E k b o m T, Dahlof B , Lanke J, Schersten B et al. Randomized trial of old and new antihypertensive drugs in elderly patients: cardiovascular mortality and morbidity the Swedish Trial in O l d Patients with Hypertension-2 study. Lancet 1999; 354(9192):1751-6.  69.  L i u L . Wang J G . G o n g L . L i u G . Staessen J A for the Systolic Hypertension in China (Syst-China) Collaborative Group. Comparison of active treatment and placebo in older Chinese patients with isolated systolic hypertension. Journal of Hypertension 1998. 16(12 Pt l):1823-9.  70.  Staessen J A , Fagard R, Thijs L , Celis FI, Arabidze G G , Birkenhager W H , et al. for The Systolic Hypertension in Europe (Syst-Eur) Trial Investigators. Randomized doubleblind comparison of placebo and active treatment for older patients with isolated systolic hypertension. Lancet 1997; 350(9080):757-64. .  71.  G r i m m R H Jr. Flack J M . Grandits G A . Elmer PJ. Neaton J D . Cuder J A . Lewis C. McDonald R. Schoenberger J. Stamler J. Long-term effects on plasma lipids of diet and drugs to treat hypertension. Treatment of M i l d Flypertension Study ( T O M H S ) Research Group. J A M A 1996; 275(20):1549-56.  248  72.  Julius S, Kjeldsen S E , Weber M . et al for the V A L U E trial group. Outcomes in hypertensive patients at high cardiovascular risk treated with regimens based on valsartan or amlodipine: the V A L U E randomised trial. Lancet 2004; 363 (9426): 20222031.  73.  Rosei E A , D a l Palu C, Leonetti G , Magnani B , Pessina A , Zanchetti A . and the V H A S Investigators. Clinical results of the Verapamil in Hypertension and Atherosclerosis Study. J Hypertens 1997; 15(11):1337-44.  74.  Cook D J , Mulrow C D , Haynes R B . Systematic reviews: synthesis of best evidence for Clinical Decisions. Annals of Internal Medicine 1997; 126: 376-380.  75.  Mulrow C D . Systematic Reviews: Rationale for systematic reviews. B M J 1994; 309: 597-599.  76.  Clarke M , Oxman A D , editors. Cochrane Reviewers' Handbook 4.1 [updated June 2000]. In: Review Manager (RevMan) [Computer program]. Version 4.1. Oxford, England: The Cochrane Collaboration, 2000.  77.  McAHster F A , Clark H D , van Walraven C, Straus S E , Lawson F M E , Moher D , Mulrow C D . The medical review article revisited: has the science improved? A n n . Intern Med. 1999; 131:947-951.  78.  The Principles of the Cochrane Collaboration. [Online]. 2006 [cited 2006 July 26]. Available from: URL:http:/ /www.cochrane.org/docs/tenprinciples.htm  79.  Jadad A R , Cook D J , Jones A , Klassen T P , Tugwell P, Moher M , Moher D . Methodology and reports of systematic reviews and meta-analyses: A comparison of Cochrane reviews with articles published in paper-based journals. J A M A 1998; 280(3): 278-180.  80.  Wong M M Y , Heran BS, Wright J M . Blood pressure lowering efficacy of calcium channel blockers for primary hypertension. The Cochrane Database of Systematic Reviews 2002, Issue 2. A r t No.: CD003657. D O I : 10.1002/14651858.CD003657.  81.  Musini V M . A Systematic Review of the Blood pressure lowering efficacy of thiazide and loop diuretics in the treatment of primary hypertension. M.Sc. Thesis, University of British Columbia 2000.  82.  Medical Research Council Working Party. M R C trial of treatment of mild hypertension: principal results. B M J 1985; 291:97-104.  83.  Boissel JP, Duperat B , Leizorovicz A . The phenomenon of regression to the mean and clinical investigation of the blood cholesterol lowering drugs. Eur. J. Clin. Pharmacol. 1980; 17: 227-230.  84.  Jadad A R . Randomised Controlled Trials. London: B M J Books, 1998.  249  85.  Heran BS, Jauca C D , Wright J M . Development of an optimal search strategy for finding trials demonstrating A C E inhibitor blood pressure lowering efficacy. 10 fnternational Cochrane Colloquium, Stavanger, Norway 2002. th  86.  W o l f F M , Guevara JP. fmputarion of missing data in systemadc reviews: so what is the standard deviation? 9 International Cochrane Colloquium, Lyon, France 2001; 1: pa007. th  87.  Moher D , Klassen T P , Jones A L , Cook D , Jadad A R , Tugwell P, Moher M . Chapter 23: Assessing the quality of reports of randomised trials included in meta-analyses: attitudes, practice, evidence and guides, fn: Stevens A , Abrams K , Brazier J, Fitzpatrick R and Lilford R, editors. The Advanced Handbook of Methods in Evidence Based Flealthcare. London: S A G E Publications; 2001. p. 409-425.  88.  Jadad A R , Moore R A , Carrol D , et al. Assessing the quality of reports of randomized clinical trials: Is blinding necessary? Contr Clin Trials 1996; 17:1-12.  89.  Fleiss J L . The statistical basis of meta-analysis. Research 1993; 2: 121-145.  90.  Sweetman SC (Ed), Martindale: The Complete Drug Pharmaceutical Press. Electronic version, (34 Edition [2006]).  Statistical Methods in Medical  Reference.  London:  th  91.  Bucher H C , Guyatt G H , Griffith L E , Walker SD. The results of direct and indirect treatment comparisons in meta-analysis of randomized controlled trials. J. Clin. Epidemiol 1997; 50(6):683-691.  92.  Song F, Altman D G , Glenny A - M , Deeks JJ. Vakdity of indirect comparison for estimating efficacy of competing interventions: empirical evidence from published meta-analyses. B M J 2003; 326:472-476.  93.  Law M R , Wald N J , Morris J K , Jordan R E . Value of low dose combination treatment with blood pressure lowering drugs: analysis of 354 randomised trials. B M J 2003; 326:1427-1434.  94.  Chrysant S G , Marbury T C , Robinson T D . Antihypertensive efficacy and safety of olmesartan medoxomil compared with amlodipine for mild-to-moderate hypertension. Journal of Ffuman Hypertension 2003; 17(6):425-32.  95.  Farsang C, Kawecka-Jaszcs K , Langan J, et al. Antihypertensive effects and tolerability of candesartan cilexetil alone and in combination with amlodipine. Clin Drug Invest 2001; 21 (1): 17-23  96.  Frick M H , McGibney D , Tyler H M . Amlodipine: a double-blind evaluation of the dose-response relationship in mild to moderate hypertension. Journal of Cardiovascular Pharmacology 1988; 12 Suppl 7:S76-S78.  250  97.  Frick M H , McGibney D , Tyler H M . A dose-response study of amlodipine in mild to moderate hypertension. Journal of Internal Medicine 1989; 225(2):101-105.  98.  Frishman W H , R A M C V , McMahon F G , Chrysant S G , Graff A , Kupiec J W et al. Comparison of amlodipine and benazepril monotherapy to amlodipine plus benazepril in patients with systemic hypertension: a randomized, double-blind, placebocontrolled, parallel-group study. The Benazepril/Amlodipine Study Group. Journal of Clinical Pharmacology 1995; 35(11):1060-1066.  99.  Kuschnir E , Acuna E , Sevilla D , Vasquez J, Bendersky M , Resk J et al. Treatment of patients with essential hypertension: amlodipine 5 mg/benazepril 20 mg compared with amlodipine 5 mg, benazepril 20 mg, and placebo. Clinical Therapeutics 1996; 18(6):1213-1224.  100. Licata G , Scaglione R, Ganguzza A , Parrinello G , Costa R, Merlino G et al. Effects of amlodipine on renal haemodynamics in mild to moderate hypertensive patients. A randomized controlled study versus placebo. European Journal of Clinical Pharmacology 1993; 45(4):307-311. 101. Mehta J L , Lopez L M , Vlachakis N D , Gradman A H , Nash D T , O'ConneU M T et al. Double-blind evaluation of the dose-response relationship of amlodipine in essential hypertension. American Heart Journal 1993; 125(6):1704-1710. 102. Mroczek WJ, Burris JF, Allenby K S . A double-blind evaluation of the effect of amlodipine on ambulatory blood pressure in hypertensive patients. Journal of Cardiovascular Pharmacology 1988; 12 Suppl 7:S79-S84. 103. Mroczek WJ, Burris JF, Klein J. A double-blind evaluation of the effect of amlodipine on ambulatory blood pressure. Postgrad M e d J 1991; 67 Suppl 5:S24-S27. 104. Mroczek WJ, Burris J F , Allenby K S . Effect of amlodipine on 24-hour ambulator}' blood pressure in hypertensive patients. Journal of Cardiovascular Pharmacology 1991; 17 Suppl. 1: S13-S15. 105. Burris J F , Allenby K S , Mroczek WJ. The effect of amlodipine on ambulator}' blood pressure in hypertensive patients. American Journal of Cardiology 1994; 73(3):39A43A. 106. Pool J, Kaihlanen P, Lewis G , Ginsberg D , Oparil S, Glazer R et al. Once-daily treatment of patients with hypertension: a placebo-controlled study of amlodipine and benazepril vs amlodipine or benazepril alone. Journal of Fluman Hypertension 2001; 15:495-498. 107. Hart W , Flolwerda N J . Barnidipine, a novel calcium antagonist for once-daily treatment of hypertension: a multicenter, double-blind, placebo-controlled, doseranging study. Dutch Barnidipine Multicenter Study Group. Cardiovascular Drugs & Therapy 1997; ll(5):637-643.  251  108. Chrysant S G , Chrysant C, Trus J, Hitchcock A . Monotherapy of hypertension with darodipine: a new calcium-channel blocker. Clinical Cardiology 1988; ll(7):467-472. 109. Burns J F , Weir M R , Oparil S, Weber M , Cady WJ, Stewart W H . A n assessment of diltiazem and hydrochlorothiazide in hypertension. Application of factorial trial design to a multicenter clinical trial of combination therapy. J A M A 1990; 263(11):1507-1512. 110. Chan P, L i n C N , Tomlinson B , L i n T H , Lee Y S . Additive effects of diltiazem and lisinopril in the treatment of elderly patients with mild-to-moderate hypertension. American Journal of Hypertension 1997; 10(7 Pt l):743-749. 111. Cushman W C , Cohen J D , Jones R P , Marbury T C , Rhoades R B , Smith L K . Comparison of the fixed combination of enalapril/diltiazem E R and their monotherapies in stage 1 to 3 essential hypertension. American Journal of Flypertension 1998; 11(1 Pt l):23-30. 112. Djian J, Ferme I, Zannad F, Chaignon M , Asmar R. Effects of sustained-release diltiazem on blood pressure and serum lipids: a mulricenter, randomized, placebocontrolled study. Journal of Cardiovascular Pharmacology 1990; 16 Suppl LS38-S45. 113. Fekcetta J V , Serfer H M , Cuder N R , Comstock TJ, Huber G L , Weir M R et al. A doseresponse trial of once-daily diltiazem. American Heart Journal 1992; 123(4 Pt 1):10221026. 114. Fiddes R, Heym H , Hilty W, Lewin A J , Codispoti J, McNally C et al. Blood pressure control with diltiazem X R , a novel extended-release formulation of diltiazem FIC1, in mature and elderly hypertensive patients. Clinical Therapeutics 1994; 16(2):209-221. 115. Herpin D , Brion N , Debregeas B. Comparison of the antihypertensive effects of sustained-release diltiazem 240 and 300 mg in patients with mild to moderate hypertension with analysis of ambulatory blood pressure profiles. Current Therapeutic Research 1990,47(2): 328-338 116. Meeves S G , Park G D . The use of confidence intervals to describe the precision of trough/peak ratios for diltiazem C D in the treatment of hypertension. Journal of Clinical Pharmacology 1994; 34(3):231-235. 117. Pool P E , Applegate W B , Woehler T, Sandall P, Cady WJ. A randomized, controlled trial comparing diltiazem, hydrochlorothiazide, and their combination in the therapy of essential hypertension. Pharmacotherapy 1993; 13(5):487-493. 118. Prisant L M . Ambulator)' blood pressure profiles in patients treated with once-daily diltiazem extended-release or indapamide alone or in combination. American Journal of Therapeutics 2000; 7(3):177-184. 119. Weir M R , Weber M A , Punzi FIA, Serfer FIM, Rosenblatt S, Cady WJ. A dose escalation trial comparing the combination of diltiazem SR and hydrochlorothiazide  252  with the monotherapies in patients with essential hypertension. Journal of Human Hypertension 1992; 6 ( 2 ) : 1 3 3 - 1 3 8 . 120. Whelton A , E f f J, Magner D J . Sustained antihypertensive activity of diltiazem SR: double-blind, placebo-controlled study with 24-hour ambulatory blood pressure monitoring. Journal of Clinical Pharmacology 1992; 32(9):808-815. 121. Black H R . Elliott W J , Weber MA, Frishman W H , Strom J A , Liebson P R et al. Oneyear study of felodipine or placebo for stage 1 isolated systolic hypertension. Hypertension 2001; 38:1118-1123. 122. Felodipine, a new calcium antagonist, as monotherapy in mild or moderate hypertension. Cooperative study group. Drugs 1987; 34 Suppl 3:139-148. 123. Hamilton D V , Barnes P C , Bowles R M , Rajaratnam D V , Comerford M B , Ginks et al. Monotherapy with felodipine, a new calcium antagonist, in mild and moderate hypertension. Journal of Cardiovascular Pharmacology 1987; 10 Suppl 10:S111-S113. 124. Fagan T O Blood pressure reduction and tolerability of felodipine E R in older and younger hypertensive patients. The Felodipine E R in the Elderly versus Young Working Group. Journal of the American Geriatrics Society .1997; 45(6):712-717. 125. Fetter M , Bastian M , Mann G , Haasis R, Diener H C , Haux R. Antihypertensive therapy does not alter cerebral bloodflow velocities under resting conditions. [German]. A k t . Neurol. 1994;  21:177-182. 126. Gradman A H , Order N R , Davis PJ, Robbins J A , Weiss RJ, W o o d B C . Combined enalapril and felodipine extended release (ER) for systemic hypertension. EnalaprilFelodipine E R Factorial Study Group. American Journal of Cardiology 1997; 79(4):431-435. 127. Kiesewetter FI, Birk A , Jung F, Radtke H . Effects of felodipine on the arterial blood pressure and fluidity of blood in patients with arterial hypertension W H O stage I. Clinical Hemorheology 1994; 14(3):355-367. 128. Liedholm H , Melander A . A placebo-controlled dose-response study of felodipine extended release in hypertensive patients. Journal of Cardiovascular Pharmacology 1989; 14(1):109-113. 129. Scholze J, Bauer B , Massaro J. Antihypertensive profiles with ascending dose combinations of ramipril and felodipine E R . Clinical & Experimental Hypertension (New York) 1999; 21(8):1447-1462. 130. van Ree J W , van der P o l G A . L o w dosages of felodipine E R once daily as monotherapy in elderly hypertensive patients: effect on ambulator}' blood pressure and quality of life. Journal of Fluman Hypertension 1996; 10(9):613-618.  253  131. Weber M A , Goldberg A I , Faison E P , Lipschutz K , Shapiro D A , Nelson E B et al. Extended-release felodipine in patients with mild to moderate hypertension. Felodipine E R Dose-Response Study Group. Clinical Pharmacology & Therapeutics 1994; 55(3):346-352. 132. Wester A . Felodipine monotherapy vs. placebo: a dose-response study in hypertension. Binational M C Study Group in U . K . and Holland. Journal of Cardiovascular Pharmacology 1990; 15 Suppl 4:S59. 133. Wester A , Lorimer A R , Westberg B . Felodipine extended release in mild to moderate hypertension. Current Medical Research & Opinion 1991; 12(5):275-281 134. Arosio E , Pancera P, Priante F, Ribul M , De Marchi S, Lechi A . Effects of sustainedrelease isradipine on blood pressure and peripheral hemodynamics in hypertensive padents. Clinical Therapeutics 1993; 15(4):705-713. 135. Burger K J , Dehner R. Zirkadiane antihypertensive wirkung von isradipin retard bei patienten mit essentieller hypertonic im vergleich zu plazebo. Orcadian antihypertensive effects of isradipine retard in patients, with essential hypertension in comparison to placebo [German]. Arzneimittel-Forschung 1993; 43(9):958-962. 136. Chrysant S G , Cohen M . Sustained blood pressure control with controlled-release isradipine. American Journal of Hypertension 1995; 8(l):87-89. 137. Chrysant S G , Cohen M . Sustained blood pressure control with controlled-release isradipine (isradipine-CR). Journal of Clinical Pharmacology 1995; 35(3):239-243. 138. Holmes D G . Isradipine: a slow-release formulation given once daily controls blood pressure for 24 h. American Journal of Hypertension 1993; 6(3 Pt 2):74S-76S. 139. Multicenter evaluation of the safety and efficacy of isradipine in hypertension. The Italian-Belgian Isradipine Study Group. American Journal of Medicine 1989; 86(4A):94-97. 140. K i r c h W , Burger K J , Weidinger G , Welzel D . Efficacy and tolerability of the new calcium antagonist isradipine in essential hypertension. Journal of Cardiovascular Pharmacology 1990; 15 Suppl 1:S55-S59. 141. Man in't Veld, V , Holmes D G , Lasance A , de Zwart P. Isradipine twice daily lowers blood pressure over 24 FI. American Journal of Ffypertension 1991; 4(2 Pt 2): 131S134S. 142. O'Grady J, Kritz H , Schmid P, Pirich C, Sinzinger FI. Effect of isradipine on in-vivo platelet function. Thrombosis Research 1997; 86(5):363-371. 143. Pittrow D B , Andsperger A , Welzel D , Wambach G , Schardt W, Weidinger G . Evaluation of the efficacy and tolerability of a low-dose combination of isradipine and  254  spirapril in the fkst-line treatment of mild to moderate essential Cardiovascular Drugs & Therapy 1997; 11(5):619-627.  hypertension.  144. Prisant L M , Carr A A . Assessment of electrocardiographic ischemia in hypertensive patients treated with isradipine or placebo. Journal of Clinical Pharmacology 1991; 31(3):233-237. 145. Youssef S, Osman L , Sabbour M S . Serum lipoprotein profile under different antihypertensive therapy. International Journal of Clinical Pharmacology Research 1992; 12(3):109-116. 146. Youssef S, Osman L , Sabbour M S . Serum lipoprotein profile under antihypertensive therapy. Cardiovascular Risk Factors 1993; 3(2): 107-111.  different  147. Rizzini P, Castello C, Salvi S, Recchia G . Efficacy and safety of lacidipine, a new longlasting calcium antagonist, in elderly hypertensive patients. Journal of Cardiovascular Pharmacology 1991; 17 Suppl 4:S38-S43. 148. Barbagallo M , Barbagallo S G . Efficacy and tolerability of lercanidipine in monotherapy in elderly patients with isolated systolic hypertension. Aging (Milano) 2000; 12(5):375379. 149. Circo A . Active dose findings for lercanidipine in a double-blind, placebo-controlled design in patients with mild to moderate hypertension. Journal of Cardiovascular Pharmacology 1997; 29(Suppl. 2):S21-S25. 150. N i n c i M A , Magliocca R, Malliani A . Efficacy and tolerability of lercanidipine in elderly patients with mild to moderate hypertension in a placebo-controlled, double-blind study. Journal of Cardiovascular Pharmacology 29(Suppl. 2):S40-S44. 151. Omboni S, Zanchetti A . Antihypertensive efficacy of lercanidipine at 2.5, 5 and 10 mg in mild to moderate essential hypertensives assessed by clinic and ambulatory blood pressure measurements. Multicenter Study Investigators. Journal of Hypertension 1998; 16(12 Pt 1):1831-1838. 152. Rimoldi E , Lumina C, Giunta L , Moscatelli P, Gatto E . Evaluation of the efficacy and tolerability of tow different formulations of lercanidipine versus placebo after oncedialy administration in mild to moderate hypertensive patients. Current Therapeutic Research 1993; 54(2):248-253. 153. Rimoldi E , Lumina C, Sega R, Libretti A . Efficacy and tolerability of lercanidipine once a day versus placebo in mild to moderate arterial hypertension. Acta Therapeutica 1994; 20:23-31. 154. Meilink-Hoedemaker L J , Pool J, Muste-Heijns M M . A double-bund placebocontrolled cross-over study with lidoflazine (Clinium) in post-infarction patients. Acta Medica Scandmavica 1976; 199(1 -2):17-23.  255  155. Fogari R, Zoppi A , Lusardi P, Mugellini A . Efficacy and tolerability of manidipine hydrochloride in the long-term treatment of mild-moderate hypertension. Manidipine Efficacy in Long-Term Treatment Group. Blood Pressure 1996; Supplement. 5:24-28. 156. Fogari R, Zoppi A , Mugellini A , Preti P, Corradi L , Lusardi P. Effect of low-dose manidipine on ambulatory blood pressure in very elderly hypertensives. Cardiovascular Drugs & Therapy 1999; 13(3):243-248. 157. Bernink PJ, Prager G , Schelling A , Kobrin I. Antihypertensive properties of the novel calcium antagonist rnibefradil (Ro 40-5967): a new generation of calcium antagonists? Mibefradil International Study Group. Hypertension 1996; 27(3 Pt l):426-432. 158. Bursztyn M , Kadr H , Tilvis R, Martina B , Oigman W , Talberg J et al. Mibefradil, a novel calcium antagonist, in elderly patients with hypertension: favorable hemodynamics and pharmacokinetics. American Heart Journal 1997; 134(2 Pt 1):238247. 159. Oparil S, Kobrin I, Abernethy D R , Levme BS, Reif M C , Shepherd A M . Doseresponse characteristics of mibefradil, a novel calcium antagonist, in the treatment of essential hypertension, [erratum appears in A m J Hypertens 1997 Sep;10(9 Pt 1): 1081]. American Journal of Hypertension t997; f 0(7 Pt l):735-742. 160. Asplund J. Nicardipine hydrochloride in essential hypertension—a controlled study. British Journal of Clinical Pharmacology 1985; 20 Suppl L120S-124S. 161. Bellet M , Pagny J Y , Chatellier G , Corvol P, Menard J. Evaluation of slow release nicardipine in essential hypertension by casual and ambulatory blood pressure measurements. Effects of acute versus chronic administration. Journal of Hypertension 1987; 5(5):599-604. f 62. Bellet B , Pagny J Y , Chatellier G , Frere R, Menard D , Corvol P et al. Evaluation en double insu de la nicardipine a liberation differee par differentes methods de measures tensionnelles. Valeur predictive de la response aigue a la nicardipine intraveineuse. [Double-blind evaluation of slow-release nicardipine using different methods of blood pressure measurement. Predictive value of the acute response to intravenous nicardipine]. [French]. Archives des Maladies du Coeur et des Vaisseaux 1987; 80(6):851-855. 163. D e Cesaris R, Ranieri G , Andriani A , Filitti V , Bonfantino M V , Lamontanara G et al. Azione antiipertensiva della nicardipina retard nelle 24 ore e suo comportamento nello stress. [Antihypertensive action of nicardipine retard in 24 hours and its effect on stress], [Italian]. Minerva Medica 1993; 84(10):533-539. 164. Fagan T C , Tyler E D , Reitman M A , Kenley S, Weber M A . Sustained-release nicardipine in mild-to-moderate hypertension. Chest 1993; 104(2):427-433. 165. Marcadet D M , Blanc A S , Lopez A A , Preziosi JP, Guignard M M , Baggio S et al. Efficacite et tolerance de la nicardipine L P 50 mg dans l'hypertension arterielle du  256  sportif. [Efficacy and tolerance of L A 50 mg nicardipine in hypertensive athletes]. [French]. Archives des Maladies du Coeur et des Vaisseaux 1991; 84(11):1569-1574. 166. Mazzola C, Borgnino C, Colombo D , Lissoni F, Serra G , Turri D R et al. Antihypertensive and hemodynamic effects of slow-release nicardipine. International Journal of Clinical Pharmacology, Therapy, & Toxicology 1988; 26(10):503-508. 167. Scuteri A , Cacciafest M , Bellucci CR, De Propris A M , D i Bernardo M G , Piccirillo G . Acute effects of long-acting nicardipine and enalapril on the quality of life in elderly patients. Current Therapeutic Research 1992; 51(5): 773-778. 168. Soro S, Cocca A , Pasanisi F, Tommaselli A , de Simone G , Costantino G et al. The effects of nicardipine on sodium and calcium metabolism in hypertensive patients: a chronic study. Journal of Clinical Pharmacology 1990; 30(2):133-137. 169. de Simone G , Costantino G , Soro S, D i Lorenzo L , Moccia D , Buonissimo S et al. Effects of nicardipine on left ventricular hemodynamic patterns in systemic hypertension. American Journal of Flypertension 1989; 2(3 Pt 1):139-145. 170. Carr A A , Bottini P B , Feig P, Prisant L M , Mulligan S, Devane J G et al. Effectiveness of once-daily monotherapy with a new nifedipine sustained release calcium antagonist. American Journal of Cardiology 1992; 69(13):28E-32E. 171.  Carr A A , Bottini P B , Feig P U , Karlos M L , Schwarz L A , Mulligan S et al. Once-daily monotherapy of hypertension with nifedipine sustained release (20 mg to 100 mg). Drug Invest. 1992; 4(2): 184-191.  172. de Simone G , Ferrara L A , di Lorenzo L , Lauria R, Fasano M . Effects of slow-release nifedipine on left ventricular mass and systolic function in mild or moderate hypertension. Current Therapeutic Research 1984; 36(3): 537-544. 173.  de Simone G , Ferrara L A , Fasano M L , D i Lorenzo L , Lauria R. Slow-release nifedipine versus placebo in the treatment of arterial hypertension. A double blind ergometric evaluation of cardiac workload. Japanese Heart Journal 1985; 26(2):219225.  174. Eggertsen R, Hansson L . Effects of treatment with nifedipine and metoprolol in essential hypertension. Eur J Clin Pharmacol 1982; 21:389-390. 175. Fadayomi M O , Akinroye K K , Ajao R O , Awosika L A . Monotherapy with nifedipine for essential hypertension in adult blacks. Journal of Cardiovascular Pharmacology 1986; 8(3):466-469. 176. Feig P U , Gibson L , Mac Carthy E P , Pettis P P , Schwartz L . The efficacy and safety of once-daily nifedipine coat-core in the treatment of mild-to-moderate hypertension. Adalat C C Cooperative Study Group, [erratum appears in Clin Ther 1994 JanFeb;16(l):125]. Clinical Therapeutics 1993; 15(6):963-975.  257  177. Ferrera L A , de Simone G , Mancini M , Fasano M L , Pasanisi F, Vallone G . Changes in left ventricular mass during a double-blind study with chlorthalidone and slow-release nifedipine. Eur J Clin Pharmacol 1984; 27:525-528. 178. Harder S, Rietbrock S, Thurmann P. Antihypertensive efficacy of a slow release nifedipine tablet formulation given once daily in patients with mild to moderate hypertension. A placebo-controlled, double-blind parallel-group trial. ArzneimittelForschung 1994; 44(2):133-136. 179. Jueng C, Flalperin A K , Flashimoto F, Callender K . Nifedipine G I T S and hydrochlorothiazide in essential hypertension. Journal of Clinical Hypertension 1987; 3(4):695-703. 180. Serradimigni A , Perchicot E , Griener L , Kerihuel J C , Benichou M . [Antihypertensive effect of a new form of nifedipine compared to a beta-blocker during chronic administration]. [French]. Archives des Maladies du Coeur et des Vaisseaux 1985; 78 Spec No:73-81. 181. Toal C B . Efficacy of a low dose nifedipine GITS (20 mg) in patients with mild to moderate hypertension. Canadian Journal of Cardiology 1997; 13(10):921-927. 182. Zachariah P K , Schwartz G L , Sheps S G , Schirger A , Carlson C A , Moore A G . Antihypertensive effects of a new sustained-release formulation of nifedipine. Journal of Clinical Pharmacology 1990; 30(11):1012-1019. 183. Zanchetti A for the Italian GITS Study Group, Bianchi L , Amigoni S, Omboni S, Villani A , Ravogli A . Antihypertensive effects of nifedipine gastrointestinal therapeutic system on clinic and ambulatory blood pressure in essential hypertensives. Journal of Hypertension 1993; 11 (Suppl 5):S334-S335. 184. Zanchetti A . Trough and peak effects of a single daily dose of nifedipine gastrointestinal therapeutic system (GITS) as assessed by ambulatory blood pressure monitoring. Italian Nifedipine G I T S Study Group. Journal of Hypertension Supplement 1994; 12(5):S23-S27. 185. Zanchetti A . The 24-hour efficacy of a new once-daily formulation of nifedipine. Italian Nifedipine GITS Study Group. Drugs 1994; 48 Suppl 1:23-30. 186. Zanchetti A for the Italian Nifedipine G I T S Study Group; Bianchi L , Bozza M , Omboni S, Villani A , Ravogli. Antihypertensive effects of nifedipine G I T S on clinic and ambulatory blood pressures in essential hypertensives. High Blood Press. 1994; 3: 45-56. 187. Hoffmann A , Kraul F l , Burkardt I. Nilvadipine in hypertension—experience in ambulatory treatment. International Journal of Clinical Pharmacology &c Therapeutics 1997; 35(5):195-203.  258  188. Weir M R , Vkchkis N D , DeQuattro V , Douglas J, Svetkey L P , Singh S et al. Evaluation of the clinical pharmacology of nilvadipine in patients with mild to moderate essential hypertension. Journal of Clinical Pharmacology 1990; 30(5):425437. 189. Opie L H , MuUer F O , Myburgh D P , Rosendorff C, Sareli P, Seedat Y K et al. for the Ambulatory Nisoldipine Coat-Core Hypertension Outpatient Response ( A N C H O R ) Investigators. Efficacy and tolerability of nisoldipine coat-core formulation in the treatment of essential hypertension: The South African Multicenter A N C H O R Study, [erratum appears in A m J Hypertens 1997 Jun;10(6):696]. American Journal of Hypertension 1997; 10(3):250-260. 190. Asmar R, Benetos A , Brahimi M , Chaouche K , Safar M . Arterial and antihypertensive effects of nitrendipine: a double-blind comparison versus placebo. Journal of Cardiovascular Pharmacology 1992; 20(6):858-863. 191. Asmar R, Benetos A , Ghodsi N , Meilhac B , Safar M . Hemodynamique arterielle regionale et circulation carotidienne dans l'hypertension arterielle essentielle. Application a la nitrendipine chez le sujet age de plus de 40 ans. [Regional arterial hemodynamics and carotid circulation in essential arterial hypertension. Effect of nitrendipine in patients over 40 years of age]. [French]. Presse Medicare 1993; 22(35):1770-1775. 192. Ferrara L A , Fasano M L , de Simone G , Soro S, Gagliardi R. Antihypertensive and cardiovascular effects of nitrendipine: a controlled study vs. placebo. Clinical Pharmacology & Therapeutics 1985; 38(4):434-438. 193. Fodor J G , Chockakngam A , Cifkova R, Strong H A , Cobby J, Mukherjee J. Efficacy of once daily nitrendipine in mild hypertension: comparison with placebo. Canadian Journal of Cardiology 1991; 7(5):223-228. 194. Gerritsen T A , Bak A A , Stolk RP, Jonker JJ, Grobbee D E . Effects of nitrendipine and enalapril on left ventricular mass in patients with non-insulin-dependent diabetes mellitus and hypertension. Journal of Hypertension 1998; 16(5):689-696. 195. Gregorio F, D i n i F L , Ambrosi F , Cristallini S, Pedetti M , Filipponi P. Efficacia antiipertensiva ed effetti sul metabolismo glucidico della nitrendipina: studio clinico controllato versus placebo. [Antihypertensive efficacy of nitrendipine and its effects on carbohydrate metabolism. A controlled clinical study versus placebo]. [Italian]. Minerva Cardioangiologica 1991; 39(4):141-148. 196. Kuschnir E , Castro R, Bendersky M , Sgammini H , Guzman L , Sgarlatta H et al. Hemodynamic effects of nitrendipine on systolic ventricular function, diastolic ventricular function, and peripheral circulation in essential hypertension. Journal of Cardiovascular Pharmacology 1988; 12 Suppl 4:S36-S44.  259  197. Lederle R M , Klaus D , Wilk S, Knaup G . Low-dose nitrendipine in mild hypertension: a double-blind, placebo-controlled, comparative study. Journal of Cardiovascular Pharmacology 1991; 18 Suppl LS48-S50. 198. Klaus D , Wilk S, Knaup G . Nitrendipin bei milder hypertonic. Eine doppelblinde, placebokontrollierte vergleichsstudie. [Nitrendipine in the treatment of mild hypertension. A double blind placebo-contolled comparative study] [German]. Zeitschrift fur Allgemeinmedizin 66(35-36): 1038-1042. 199. Maclean.D, Mitchell E T , Lewis R, Irvine N , McLay JS, M c E w e n J et al. Comparison of once daily atenolol, nitrendipine and their combination in mild to moderate essential hypertension. British Journal of Clinical Pharmacology 1990; 29(4):455-463. 200. Paolisso G , Aceto E , Cennamo G , D'Amore A , Varricchio M . Metabolic effects of nitrendipine. Clinical Therapeutics 1991; 13(6):695-698. 201. Roca-Cusachs A , Torres F, Floras M , Rios J, Calvo G , Delgadillo J, et al. Nitrendipine and enalapril combination therapy in mild to moderate hypertension: assessment of dose-response relationship by a clinical trial of factorial design. Journal of Cardiovascular Pharmacology 2001; 38:840-849. 202. Rosenthal J, Flittel N , Stumpe K O . Pranidipine, a novel calcium antagonist, once daily, for the treatment of hypertension: a multicenter, double-blind, placebo-controlled dose-finding study. Cardiovascular Drugs & Therapy 1996; 10(l):59-66. 203. Blanchett D G , Corder C N , Kezdi P, Margolis R, Jain A , Ryan JR et al. Antihypertensive effect of tiapamil from ambulatory and clinic methods. International Journal of Clinical Pharmacology, Therapy, & Toxicology 1991; 29(2):59-63. 204. Carr A A , Bottini P B , Pnsant L M , Fisher L D , Devane J G , O'Brien D E et al. Oncedaily verapamil in the treatment of mild-to-moderate hypertension: a double-blind placebo-controlled dose-ranging study. Journal of Clinical Pharmacology 1991; 31(2):144-150 :  205. DeQuattro V , Lee D . Fixed-dose combination therapy with trandolapril and verapamil SR is effective in primary hypertension. Trandolapril Study Group. American Journal of Flypertension 1997; 10(7 Pt 2):138S-145S. 206. DeQuattro V , Lee D , Messerli F. for the Trandolapril Study Group. Efficacy of combination therapy with trandolapril and verapamil sr in primary hypertension: a 4 x 4 trial design. Clinical & Experimental Hypertension (New York) 1997; 19(3):373-387. 207. Levine J H , Applegate W B , and the Trandolapril/Verapamil Study Group. Trandolapril and Verapamil slow release in the treatment of hypertension: a doseresponse assessment with the use of a multifactorial trial design. Current Therapeutic Research 1997; 58(6):361-374.  260  208. Levine J H , Ferdinand K C , Cargo P, Laine H , Lefkowitz M . Additive effects of verapamil and enalapril in the treatment of mild to moderate hypertension. American Journal of Hypertension 1995; 8(5 Pt l):494-499. 209. McMahon F G , Reder R F . The relationship of dose to the antihypertensive response of verapamil-sustained release in patients with mild to moderate essential hypertension. The Verapamil-SR Study Group. Journal of Clinical Pharmacology 1989; 29(11):10031007. 210. Messerli F, Frishman W H , Elliott WJ. Effects of verapamil and trandolapril in the treatment of hypertension. Trandolapril Study Group. American Journal of Flypertension 1998; 11(3 Pt l):322-327. 211. Neutel J M , Alderman M , Anders RJ, Weber M A . Novel delivery system for verapamil designed to achieve maximal blood pressure control during the early morning. American Heart Journal 1996; 132(6):1202-1206. 212. Neutel J M , Smith D G H , Weber M A . The use of chronotherapeutics to achieve maximal blood pressure reduction during the early morning blood pressure surge. C V R & R 1999:575-581. 213. Scholze J, Zilles P, Compagnone D . Verapamil SR and trandolapril combination therapy in hypertension—a clinical trial of factorial design. German Hypertension Study Group. British Journal of Clinical Pharmacology 1998; 45(5):491-495. 214. Smith D H , Neutel J M , Weber M A . A new chronotherapeutic oral drug absorption system for verapamil optimizes blood pressure control in the morning. American Journal of Hypertension 200f; 14(1):14-19. 215. Effects of verapamil SR, trandolapril, and their fixed combination on 24-h blood pressure: the Veratran Study, [erratum appears in A m J Hypertens 1998 A u g ; l l ( 8 Pt 1):1039]. American Journal of Ffypertension 1997; 10(5 Pt l):492-499. 216. von Manteuffel G E , Rakette S, Woll E M , Reinfrank J, Schiemann J. [Effectiveness and tolerance of combined verapamil retard and hydrochlorothiazide. Results of a double-blind, randomised study]. [German]. Fortschritte der Medizin 1995; 113(26):374-378. 217. White W B , Anders RJ, Maclntyre J M , Black H R , Ska D A . Nocturnal dosing of a novel delivery system of verapamil for systemic hypertension. Verapamil Study Group. American Journal of Cardiology 1995; 76(5):375-380. 218. White W B , Sica D A , Calhoun D , Mansoor G A , Anders RJ. Preventing increases in early-morning blood pressure, heart rate, and the rate-pressure product with controllec onset extended release verapamil at bedtime versus enalapril, losartan, and placebo on arising. A m Heart J 2002; 144:657-665.  261  219. Bakris G , Sica D , Ram V , Fagan T, Vaitkus P T , Anders RJ. A comparative trial of controlled-onset, extended-release verapamil, enalapril, and losartan on blood pressure and heart rate changes. AJFI 2002; 15:53-57. 220.  Glasser SP, Neutel J M , Gana TJ, Albert K S . Efficacy and safety of a once daily graded-release diltiazem formulation in essential hypertension. A J H 2003; 16:51-58.  221.  Abadie E , Villette J M , Gauville C, Tabuteau F, Fiet J, Passa P. [Effects of nifedipine on carbohydrate metabolism in the non-insulin dependent diabetic]. [French]. Diabete et Metabolisme 1985; 11 (3):141-146.  222. Ahmed J H , Elliott H L , Flosie J, Farish E , Reid J L . Effects of nicardipine on the metabolic responses to food and exercise. Journal of Fluman Hypertension 1992; 6(2):139-144. 223. Allikmets K , Parik T, Teesalu R. Antihypertensive and renal effects of isradipine in essential hypertension: focus on renin system activity. Angiology 1997; 48(ll):977-983. 224. Andersson O K , Persson B , Hedner T, Aurell M , Wysocki M . Blood pressure control and haemodynamic adaptation with the dihydropyridine calcium antagonist isradipine: a controlled study in middle-aged hypertensive men. Journal of Hypertension 1989; 7(6):465-469. 225. Andersson O K , Persson B , Widgren R, Wysocki M . Central hemodynamics an dbrachial artery compliance during therapy with isradipine, a new calcium antagonist. Journal of Cardiovascular Pharmacology 1990; 15(Suppl. 1):S87-S89. 226. Andren L , Flansson L , O r o L , Ryman T. Experience with nitrendipine—a new calcium antagonist—in hypertension. Journal of Cardiovascular Pharmacology 1982; 4 Suppl 3:S387-S391. 227. Arita M , Hashizume T, Tanigawa K , Yamamoto FI, Nishio I. A new Ca-antagonist, azelnidipine, reduced blood pressure during exercise without augmentation of sympathetic nervous system in essential hypertension: a randomized, double-blind, placebo-controlled trial. Journal of Cardiovascular Pharmacology 1999; 33(2):186-192. 228. Arzilli F, Gandolfi E , D e l Prato C, Innocenti P, Ponzanelli F, Caiazza et al. Antihypertensive effect of once daily sustained release isradipine: a placebo controlled cross-over study. European Journal of Clinical Pharmacology 1993; 44(l):23-25. 229. Bainbridge A D , Macfadyen RJ, Stark S, Lees K R , Reid J L . The antihypertensive efficacy and tolerability of a low dose combination of ramipril and felodipine E R in mild to moderate essential hypertension. Br. J Clin Pharmac 1993; 36:323-330. 230. Baylac-Domengetroy F, Baylac-Domengetroy J, Chaix A F , Elbaz-Rostykus C, Barraine R. [Relations between antihypertensive and vascular effects oj nitrendipine]. [French]. Archives des Maladies du Coeur et des Vaisseaux 1990; 83(8):1295-1299.  262  231. Bellet M , Loria Y , Lallemand A . First-step treatment of mild to moderate uncomplicated essential hypertension by a new calcium antagonist: nicardipine. Journal of Cardiovascular Pharmacology 1985; 7(6): 1149-1153. 232. Bossini A , D i Veroli C, Cavallotti G , Cagli V . Felodipine E R formulation in the treatment of mild hypertension: efficacy and tolerability vs placebo. British Journal of • Clinical Pharmacology 1990; 30(4):567-571. . 233. The British Isradipine Flypertension Group. Evaluation of the safety and efficacy of isradipine in elderly patients with essential hypertension. American Journal of Medicine 1989; 86(4A):110-114. 234. Campbell L M , Cowen K J , Cranfield FR, Goves JR, Jones D F , Lees C T et al. Felodipine-ER once daily as monotherapy in hypertension. Journal of Cardiovascular Pharmacology 1990; 15(4):569-573. 235. Capewell S, Collier A , Matthews D , Flajducka C, Collier R, Clarke B F et al. A trial of the calcium antagonist felodipine in hypertensive type 2 diabetic patients. Diabetic Medicine 1989; 6(9):809-812. 236. Carr A A , Prisant L M . The new calcium antagonist isradipine. Effect on blood pressure and the left ventricle in black hypertensive patients. American Journal of Hypertension 1990; 3(1):8-15. 237. Chalmers J, Wing L , Taylor S. Analysis of time trends, individual subject responses and background variation in crossover factorial studies with antihypertensive drugs. Journal of Hypertension - Supplement 1990; 8(4):S27-S35. 238. Clement D L , De Pue N Y , Packet L . Effect of calcium antagonists on ambulatory blood pressure and its variations. Journal of Cardiovascular Pharmacology 1987; 10 Suppl 10:S117-S119. 239. Cleroux J, Yardley C, Marshall A , Coulombe D , Lacourciere Y . Antihypertensive and hemodynamic effects of calcium channel blockade with isradipine after acute exercise. American Journal of Flypertension 1992; 5(2):84-87. 240. Cleroux J, Beaulieu M , Kouame N , Lacourciere Y . Comparative effects of quinapril, atenolol, and verapamil on blood pressure and forearm hemodynamics during handgrip exercise. American Journal of Hypertension 1994; 7(6):566-570. 241. Cox JP, O'Boyle C A , Mee F, Kelly J, Atkins N , Coakley D et al. The antihypertensive efficacy of verapamil in the elderly evaluated by ambulatory blood pressure measurement. Journal of Fluman Hypertension 1988; 2(l):41-47. 242. Cox JP, Ryan J, O'Brien E , O'Malley K . The effect of slow-release nicardipine on ambulator)' and clinic blood pressure in mild hypertension. British Journal of Clinical Pharmacology 1989; 28(l):79-82.  263  243. Crozier I G , Ikram H , Nicholls M G , L o w CJ. Twenty-four hour profile of the hypotensive action of felodipine in essential hypertension. Cardiovascular Drugs & Therapy 1990; 4(2):439-442. 244. de Bruijn B , Cocco G , Tyler H M . Multicenter placebo-controlled comparison of amlodipine and atenolol in mild to moderate hypertension. Journal of Cardiovascular Pharmacology 1988; 12 Suppl 7:S107-S109. 245. Diemont W L , Stegeman CJ, Beekman J , Siegers A M , Hagels G , de Bruijn J H . L o w dose isradipine once daily effectively controls 24-h blood pressure in essential hypertension. American Journal of Hypertension 1991; 4(2 Pt 2):163S-167S. 246. Dittrich H C , Adler J, Ong J, Reitman M , Weber M , Ziegler M . Effects of sustainedrelease nicardipine on regression of left ventricular hypertrophy in systemic hypertension. American Journal of Cardiology 1992; 69(19):1559-1564. 247. Draaijer P, K o o l M J , V a n Bortel L M , Nieman F, de Leeuw P W , van Hooff JP et al. Vascular compliance in sodium-sensitive and sodium-resistant borderline hypertensive patients. Kidney International 1995; 47(1):169-176. 248. Dupont A G , Coupez J M , Jensen P, Coupez-Lopinot R, Schoors D F , Plermanns et al. Twenty-four hour ambulator)' blood pressure profile of a new slow-release formulation of diltiazem in mild to moderate hypertension. Cardiovascular Drugs & Therapy 1991; 5(4):701-707. 249. Duprez D , D e Backer T, D e Pue N , Ffermans L , De Buyzere M , Clement D L . Effects of isradipine on peripheral hemodynamic reflex responses in mild-to-moderate essential hypertension. American Journal of Hypertension 1991; 4(2 Pt 2):194S-196S. 250. Durel L A , Hayashi PJ, Weidler D J , Schneiderman N . Effectiveness of antihypertensive medications in office and ambulatory settings: a placebo-controlled comparison of atenolol, metoprolol, chlorthalidone, verapamil, and an atenolol-chlorthalidone combination. Journal of Clinical Pharmacology 1992; 32(6):564-570. 251. Faguer de Moustier B , Paoli V , Tchobroutsky G . Metabolic controlled trial of nicardipine in type 2 diabetic patients with slight hypertension. Current Therapuetic Research 1989; 45(4): 690-704. 252. Faguer de Moustier B , Paoli V . The influence of nicardipine in type 2 diabetic patients with slight hypertension. Journal of Cardiovascular Pharmacology 1990; 16 Suppl 2:S26-S33. 253. Ferreira-Filho SR, Dorneles H . Determination of systemic haemodynamic alterations induced by slow-release nifedipine in elderly hypertensive patients using a non-invasive method: double-blind cross-sectional random study. Journal of Human Hypertension 1995;9(7):535-540.  264  254. Fogari R, Malacco E , Tettamanti F, Gnemmi A E , Milani M . Evening vs morning isradipine sustained release in essential hypertension: a double-blind study with 24 h ambulatory monitoring. British Journal of Clinical Pharmacology 1993; 35(l):51-54. 255. Forette F, Bellet M , Henry J F , Hervy M P , Poyard-Salmeron C, Bouchacourt et al. [Treatment of arterial hypertension in the aged with a calcium antagonist: nicardipine]. [French]. Archives des Maladies du Coeur et des Vaisseaux 1984; 77(11):1242-1246. 256. Forette F, Bellet M , Henry JF, Hervy M P , Poyard-Salmeron C, Bouchacourt et al. Effect of nicardipine in elderly hypertensive patients. British Journal of Clinical Pharmacology 1985; 20 Suppl 1:125S-129S. 257. Frishman WFI, Brobyn R, Brown R D , Johnson B F , Reeves R L , Wombolt D G . A randomized placebo-controlled comparison of amlodipine and atenolol in mild to moderate systemic hypertension. Journal of Cardiovascular Pharmacology 1988; 12 Suppl 7:S103-S106. 258. Frishman WFI, Brobyn R, Brown R D , Johnson B F , Reeves R L , Wombolt D G . Amlodipine versus atenolol in essential hypertension. American Journal of Cardiology 1994; 73(3):50A-54A. 259. Johnson B F , Frishman W H , Brobyn R, Brown R D , Reeves R L , Wombolt D G . A randomized, placebo-controlled, double-blind comparison of amlodipine and atenolol in patients with essential hypertension. American Journal of Hypertension 1992; 5(10):727-732. 260. Johnson B F , Frishman W H , Brobyn R, Brown R D , Reeves R L , Wombolt D G . A randomized, placebo-controlled, double-blind comparison of amlodipine and atenolol in patients with essential hypertension. Cardiovascular Reviews & Reports 1995; 16(9):21-27+31. 261. Gavras I, Mulinari R, Gavras H , Higgins JT, Reeves R L , Zawada E T et al. Antihypertensive effectiveness of the nifedipine gastrointestinal therapeutic system. American Journal of Medicine 1987; 83(6B):20-23. 262. Gebara O C , Jimenez A H , McKenna C, Mitdeman M A , X u P, Lipinska I et al. Stressinduced hemodynamic and hemostatic changes in patients with systemic hypertension: effect of verapamil. Clinical Cardiology 1996; 19(3):205-211. 263. G r i m m RFI Jr, Black H , Rowen R, Lewin A , Shi H , Ghadanfar M et al. Amlodipine versus chlorthalidone versus placebo in the treatment of stage I isolated systolic hypertension. A m J Hypertension 2002; 15(1 Pt l):31-6. 264. Hamilton B P . Treatment of essential hypertension with P N 200-110 (isradipine). American Journal of Cardiology 1987; 59(3):141B-145B.  265  265.  Harrington K , Fitzgerald P, O'Donnell P, H i l l K W , O'Brien E , O'Malley K . Short and long term treatment of essential hypertension with felodipine as monotherapy. Drugs 1987; 34(Suppl. 3 ) : 1 7 8 - 1 8 5 .  266. Hedback B , Hermann LS. Antihypertensive effect of verapamil in patients with newly discovered mild to moderate essential hypertension. Acta Medica Scandinavica Supplementum 1984; 681:129-135. 267. Herrera C R , Lewin A , Fiddes R, Friedman J, Linn W , Baker T et al. Long-acting diltiazem C D is safe and effective in a hypertensive Mexican-American population. Pharmacotherapy 1997; 17(6):1254-1259. 268. Honorato J, Azanza J, Guindo N , Suarez JR. Double-blind placebo-controlled study of the antihypertensive efficacy of nitrendipine in patients submitted to ambulatory blood pressure monitoring. Current Therapeutic Research, Clinical & Experimental 1989;46(5):932-939. 269.  Hosie J, Mulder A W . Antihypertensive effects of felodipine once daily given as monotherapy to elderly patients: a placebo comparison. Binational M C Study Group (U.K. and Holland). Journal of Cardiovascular Pharmacology 1990; 15 Suppl 4:S90.  270. Hosie J, Mulder A W , Westberg B . Felodipine once daily in elderly hypertensives. Binational M C Study Group (United Kingdom and Netherlands). Journal of Human Hypertension 1991; 5 ( l ) : 4 9 - 5 3 .  271.  Jeffrey R F , Capewell S, Brown J, Collier A , Flajducka C, Lee M R . Effects of felodipine on atrial natriuretic peptide in hypertensive non-insulin dependent diabetes mellitus. British Journal of Clinical Pharmacology 1990; 30(3):481-484.  272.  Kelemen M H , Effron M B , Valenti SA, Stewart K J . Exercise training combined with antihypertensive drug therapy. Effects on lipids, blood pressure, and left ventricular mass. J A M A 1990; 263(20):2766-2771.  273.  Stewart K J , Effron M B , Valenti SA, Kelemen M H . Effects of diltiazem or propranolol during exercise training of hypertensive men. Medicine & Science in Sports & Exercise 1990; 22(2):171-177.  274. Khalil-Manesh F, Venkataraman K , Samant D R , Gadgil U G . Effects of diltiazem on cation transport across erythrocyte membranes of hypertensive humans. Hypertension 1987; 9(l):18-23. 275.  Kjellstrom T, Blychert E , Lindgarde F. Felodipine in the treatment of hypertensive type f i diabetics: effect on glucose homeostasis. Journal of fnternal Medicine 1991; 229(3):233-239.  276. Klauser R, Speiser P, Gisinger C, Schernthaner G , Prager R. Platelet aggregation and metabolic control are not affected by calcium antagonist treatment in type II diabetes mellitus. Journal of Cardiovascular Pharmacology 1990; 15 Suppl LS93-S96.  266  277. Klauser R, Prager R, Gaube S, Gisinger C, Schnack C, Kuenburg E et al. Metabolic effects of isradipine versus hydrochlorothiazide in diabetes mellitus. [erratum appears in Hypertension 1991 May;17(5):722]. Hypertension 1991; 17(1):15-21. 278. Krakoff L R . Nicardipine monotherapy in ambulatory hypertension. American Heart Journal 1989; 117(1): 250-255.  elderly patients with  279. Lacourciere Y , Pokier L , Lefebvre J, Burford R G . Clinical efficacy of force titrated doses of diltiazem extended-release. A placebo controlled study. American Journal of Hypertension 1995; 8(3):282-286. 280. Lacourciere Y , Lenis J, Orchard R, Lewanczuk R, Houde M , Pesant Y et al. A comparison of the efficacies and duration of action of the angiotensin II receptor blockers telmisartan and amlodipine. Blood Pressure Monitoring 1998; 3(5):295-302. 281. Landmark K , Dale J. Antihypertensive, haemodynamic and metabolic effects of nifedipine slow-release tablets in elderly patients. Acta Medica Scandinavica 1985; 218(4):389-396. 282. Lessem J, Gershan R M , Madden R, Meacham J, Schumacher BJ, Sdt V B et al. Nicardipine in essential hypertension. Clinical Trials Journal 1990:27(5): 313-326. 283. Letzel EI, Bluemner E . Dose-response curves in antihypertensive combination therapy: results of a controlled clinical trial. Journal of Flypertension 1990; 8(Suppl 4): S83-S86. 284. Lewis G R , Morley K D , Lewis B M , Bones PJ. The treatment of hypertension with verapamil. New Zealand Medical Journal 1978; 87(612):351-354. 285. L o k H . Felodipine in elderly hypertensives. Dutch G P Multicentre Study Group. Journal of Fluman Hypertension 1989; 3(6):467-470. 286. Yodfat Y , Cristal N . A multicenter, double-blind, randomized, placebo-controlled study of isradipine and methyldopa as monotherapy or in combination with captopril in the treatment of hypertension. The L O M I R - M C T - I H Research Group. American Journal of Flypertension 1993; 6(3 Pt 2):57S-61S. 287. Yodfat Y , Bar-On D , Amir M , Cristal N . Quality of life in normotensives compared to hypertensive men treated with isradipine or methyldopa as monotherapy or in combination with captopril: the L O M I R - M C T - I L study. Journal of Fluman Hypertension 1996; 10(2):117-122. 288. Lorimer A R , Smedsrud T, Walker P, Tyler H M . Comparison of amlodipine and verapamil in the treatment of mild to moderate hypertension. Journal of Cardiovascular Pharmacology 1988; 12 Suppl 7:S89-S93.  267  289. Lorimer A R , Smedsrud T, Walker P, Tyler H M . A comparison of amlodipine, verapamil and placebo in the treatment of mild to moderate hypertension. Amlodipine Study Group. Journal of Human Hypertension 1989; 3(3):191-196. 290. L o w CJS, Foy S G , Chaudhary H , Crozier I G , Baskaranathan S, Ikram H . Twenty-four hour profile of the anti-hypertensive action of isradiine in essential hypertension. Blood Pressure 1993;2: 59-61. 291. Lyons D , Webster J, Benjamin N . The effect of antihypertensive therapy on responsiveness to local intra-arterial NG-monomethyl-L-arginine in patients with essential hypertension. Journal of Pfypertension 1994; 12(9):1047-1052. 292. Macphee G J A , Dutton M , Lennox I M , Williams B O . Low-dose nifedipine retard in elderly hypertensive patients. Journal of Clinical & Experimental Gerontology 1989; 11(3-4):115-129. 293. Mancia G , Omboni S, Ravogli A , Frattola A , Villani A . Ambulatory blood pressure monitoring. Drugs 1992; 44 Suppl 1:17-22. 294. Massie B M , Der E , Herman TS, Topolski P, Park G D , Stewart W H . 24-hour efficacy of once-daily diltiazem in essential hypertension. Clinical Cardiology 1992; 15(5):365368. 295. Mattarei M . De Venuto G . Biasion T. et al. Nifedipine vs prazosin in the treatment of slight-to-moderate essential arterious hypertension: Effectiveness, tolerance, and metabolic effects. Current Therapeutic Research, CUnical & Experimental 1987; 41(3): 356-366. 296. Mclnnes G T , Findlay I N , Murray G , Cleland J G , Dargie HJ. Cardiovascular responses to verapamil and propranolol in hypertensive patients. Journal of Hypertension Supplement 1985; 3 Suppl 3:S219-S221. 297. Dargie H , Cleland J, Findlay I, Murray G , Mclnnes G . Combination of verapamil and beta blockers in systemic hypertension. A m } cardiol 1986; 57:80D-82D. 298. Megnien J L , Levenson J, D e l Pino M , Simon A . Amlodipine induces a flow and pressure-independent vasoactive effect on the brachial artery in hypertension. British Journal of Clinical Pharmacology 1995; 39(6):641-649. 299. Midtbo K A , Hals O , Lauve O . A new sustained-release formulation of verapamil in the treatment of hypertension. Journal of Clinical Flypertension 1986; 2(3 Suppl):125S132S. 300. Storstein L . Midtbo K . Hals O . Myhre E . Antihypertensive effect of verapamil in relation to plasma concentrations of verapamil and its active metabolite norverapamil. Current Therapeutic Research, Clinical & Experimental 1981 ;29(1 12-119.  268  301. Morgan T O , Anderson A I E , Maclnnis RJ. A C E inhibitors, beta-blockers, calcium blockers, and diuretics for the control of systolic hypertension. American Journal of Hypertension 2001; 14(3):241-247. 302. Morgan T, Anderson A . A comparison of candesartan, felodipine, and their combination in the treatment of elderly patients with systolic hypertension. A m J Hypertens. 2002;15(6):544-9. 303. Moser M , Lunn J, Nash D T , Burris J F , Winer N , Simon G . Nitrendipine in the treatment of mild to moderate hypertension. Journal of Cardiovascular Pharmacology. 1984; 6: S1085-1089. 304. Nelson E B , Pool J L , Taylor A A . Antihypertensive activity of isradipine in humans: a new dihydropyridine calcium channel antagonist. Clinical Pharmacology & Therapeutics 1986; 40(6):694-697. 305. Neutel J M , Smith D H , Lefkowitz M P , Cargo P, Alemayehu D , Weber M A . Hypertension in the elderly: 24 h ambulator)' blood pressure results from a placebocontrolled trial. Journal of Human Flypertension 1995; 9(9):723-727. 306. Nikkila M T , Inkovaara J A , Heikkinen JT, Olsson SO. Antihypertensive effect of diltiazem in a slow-release formulation for mild to moderate essential hypertension. American Journal of Cardiology 1989; 63(17):1227-1230. 307. Nikolova K . Treatment of hypertensive venous leg ulcers with nifedipine. Methods & Findings in Experimental & Clinical Pharmacology 1995; 17(8):545-549. 308. Nilsson P, Lindholm L H , Fledner T for the D D D Study Group. The diltiazem different doses study — a dose-response study of once-daily diltiazem therapy for hypertension. J Cardiovasc Pharmacol 1996; 27(4): 469-475. 309. Ollivier JP, Wajman A , Stalla-Bourdillon A . Antihypertensive efficacy of optimally titrated doses of once-daily sustained-release diltiazem: a placebo-controlled trial. D I L P L A C O M P Study Group. Cardiology 1995; 86(6):481-487. 310. Pacheco JP, Fan F, Wright R A , Corkadel L K , Horwitz L D , Reeves R L . Monotherapy of mild hypertension with nifedipine. American Journal of Medicine 1986; 81(6A):2024. 311. Pandita-Gunawardena N D , Clarke SE. Amlodipine lowers blood pressure without affecting cerebral blood flow as measured by single photon emission computed tomography in elderly hypertensive subjects. Age & Ageing 1999; 28(5):451-457. 312. Pannarale G , Puddu P E , Monti F, Irace L , Bentivoglio M , Collauto F et al. Twentyfour-hour antihypertensive efficacy of felodipine 10 mg extended-release: the Italian inter-university study. Journal of Cardiovascular Pharmacology 1996; 27(2):255-261.  269  313. Persson B , Wysocki M , Andersson O K . Long-term renal effects of isradipine, a calcium entry blocker, in essential hypertension. Journal of Cardiovascular Pharmacology 1989; 14(l):22-24. 314. Persson B , Andersson O K , Wysocki M , Hedner T, Aurell M . Renal and hemodynamic effects of isradipine in essential hypertension. American Journal of Medicine 1989; 86(4A):60-64. 315. Persson B , Andersson O K , Wysocki M , Pledner T, Karlberg B. Calcium antagonism in essential hypertension: effect on renal haemodynamics and microalbuminuria. Journal of Internal Medicine 1992; 231(3):247-252. 316. Pool P E , Seagren SC, Salel A F . Effects of diltiazem on serum lipids, exercise performance and blood pressure: randomized, double-blind, placebo-controlled evaluation for systemic hypertension. American Journal of Cardiology 1985; 56(16):86H-91H. 317. Pool P E , Massie B M , Venkataraman K , Hirsch A T , Samant D R , Seagren SC et al. Diltiazem as monotherapy for systemic hypertension: a multicenter, randomized, placebo-controlled trial. American Journal of Cardiology 1986; 57(4):212-217. 318. Pool P E , Herron J M , Rosenblatt S, Reeves R L , Nappi J M , Staker L V et al. Metabolic effects of antihypertensive therapy with a calcium antagonist. American Journal of Cardiology 1988; 62(11):109G-113G. 319. Pool P E , Herron J M , Rosenblatt S, Reeves R L , Nappi J M , Staker L V et al. Sustainedrelease diltiazem: duration of antihypertensive effect. Journal of Clinical Pharmacology 1989; 29(6):533-537. 320. Pool P E , Nappi J M , Weber M A . Antihypertensive monotherapy with tablet (promptrelease) diltiazem: multicenter controlled trials. Cardiovascular Drugs & Therapy 1990; 4(4):1089-1096. 321. Pnsant L M , Neutel J M , Papademetriou V , DeQuattro V , Flail W D , Weir M R . L o w dose combination treatment for hypertension versus single-drug treatmentbisoprolol/hydrochlorothiazide versus amlodipine, enalapril, and placebo: combined analysis of comparative studies. American Journal of Therapeutics 1998; 5(5):313-321. 322. Prisant L M , Neutel J M , Ferdinand K , Papademetriou V , DeQuattro V , Hall et al. Low-dose combination therapy as first-line hypertension treatment for blacks and nonblacks. Journal of the National Medical Association 1999; 91(l):40-48. 323. Ramirez L C , Koffler M , Arauz C, Schnurr-Breen L , Raskin P. Effect of nifedipine GfTS on blood pressure, glucose metabolism, and lipid levels in hypertensive patients. Current Therapeutic Research, Clinical & Experimental 1992; 52(3):468-477. 324. Ricciardelli B , Argenziano L , Morisco C, Monti F , Barbato E , Trimarco B . Effects of nitrendipine on plasma levels of insulin and glucose in patients with essential  270  hypertension. Current Therapeutic Research, Clinical & Experimental 1997; 58(3): 180186. 325. Romero-Vecchione E , Vasquez J, Velasco M , Morales E , Davoli M A , Arocha et al. [Increase of renal dopamine production induced by nifedipine in hypertensive patients. Double blind vs placebo study]. [Spanish]. Archivos del Instituto de Cardiologia de Mexico 1995; 65(6):535-540. 326. Rossi L , Costa B , Tomei R, Franceschini L , Castello C, Carbonieri, Zardini P. Antihypertensive effects of lacidipine during effort in mild to moderate hypertension. J Cardtovasc Pharmacol 2002;40(2):315-21. 327. Salvetti A , Innocenti P F , Iardella M , Pambianco F, Saba G C , Rossetti M et al. Captopril and nifedipine interactions in the treatment of essential hypertensives: a crossover study. Journal of Hypertension - Supplement 1987; 5(4):S139-S142. 328. Salvetti A , Cardellino G , Pesenti M , Caiazza A , Ghisoni F, D e l Prato C et al. Antihypertensive effect of slow-release nicardipine. A placebo-controlled cross-over study. European Journal of Clinical Pharmacology 1989; 36(5):439-442. 329. Salvetti A , Magagna A , Innocenti P, Cagianelli A , Cipriani M , Gandolft et al. Chlorthalidone does not increase the hypotensive effect of nifedipine in essential hypertensives: a crossover multicentre study, [erratum appears in J Hypertens Suppl 1990 Apr;8(4):399]. Journal of Flypertension - Supplement 1989; 7(6):S250-S251. 330. Salvetti A , Magagna A , Innocenti P, Ponzanelli F, Cagianelli A , Cipriani et al. The combination of chlorthalidone with nifedipine does not exert an additive antihypertensive effect in essential hypertensives: a crossover multicenter study. Journal of Cardiovascular Pharmacology 1991; 17(2):332-335. 331. Salvetti A , Virdis A , Taddei S, Ambrosoli S, Caiazza A , Gandolfi E et al. Trough:peak ration of nifedipine gastrointestinal therapeutic system and nifedipine retard in essential hypertensive patients: an ftalian multicentre study. Journal of Hypertension 1996; 14: 661-667. 332. Schuster FI, Toka O , Toka H R , Busjahn A , Oztekin O , Wienker T F et al. A cross-over medication trial for patients with autosomal-dominant hypertension with brachydactyly. Kidney International 1998; 53(1):167-172. 333. Siche JP, Baguet JP, Fagret D , Tremel F , de Gaudemaris R, Mallion J M . Effects of amlodipine on baroreflex and sympathetic nervous system activity in mild-to-moderate hypertension. A m J Hypertens. 2001; 14(5 Pt l):424-8. 334. Simon G , Snyder D K . Altered pressor responses in long-term nitrendipine treatment. Clinical Pharmacology & Therapeutics 1984; 36(3):315-319. 335. Slonim A , Paran E , Cristal N . Effect of isradipine on factors affecting blood viscosity. American Journal of Hypertension 1991; 4(2 Pt 2):172S-174S.  271  336. Spence J D , Munoz C, Huff M W , Tokmakjian S. Effect of amlodipine on hemodynamic an dendocrine responses to mental stress. A J H 2000; 13:518-522. 337. Spieker C. Efficacy and tolerability of once-daily barnidipine in the clinical management of patients with mild to moderate essential hypertension. Blood Pressure 1998; 7 (Suppl 1): 15-21. 338. Smilde J G . The long-term efficacy and safety profile of barnidipine. Journal of Clinical Practice 2000. Supplement 114: 20-26.  International  339. Spritzer N , Medeiros C M , Spritzer TS, Medeiros N K . [24-hour antihypertensive efficacy of nitrendipine]. [Portuguese]. Arquivos Brasileiros de Cardiologia 1990; 55(l):71-74. 340. Staessen J, Lijnen P, Fagard R, Hespel P, Tan W P , Devos P et al. Effects of the new calcium entry blocker isradipine ( P N 200-110) in essential hypertension. Journal of Cardiovascular Pharmacology 1989; 13(2):271-276. ;  341. Stornello M , Valvo E V , Scapellato L . Hemodynamic, renal, and humoral effects of the calcium entry blocker nicardipine and converting enzyme inhibitor captopril in hypertensive type II diabetic patients with nephropathy. Journal of Cardiovascular Pharmacology 1989; 14(6):851-855. 342. Stornello M , Valvo E V , Scapellato L . Effect of sustained-release nicardipine, chlorthalidone, and the two drugs combined in patients with mild to moderate hypertension. Current Therapeutic Research, Clinical & Experimental 1990; 47(2):405411. 343. Suzuki M , Kanazawa A , Hasegawa M , Harano Y . Improvement of insulin resistance in essential hypertension by long-acting Ca antagonist benidipine. Clinical & Experimental Hypertension (New York) 1999; 21(8):1327-1344. 344. Szlachcic J, Tubau J F , Vollmer C, Massie B M . Effect of diltiazem on left ventricular mass and diastolic filling in mild to moderate hypertension, [see comments]. American Journal of Cardiology 1989; 63(3):198-201. 345. Tonkin A L , Wing L M , Russell A E , West M J , Bune A J , Morris M J et al. Diltiazem and atenolol in essential hypertension: additivity of effects on blood pressure and cardiac conduction with combination therapy. Journal of Hypertension 1990; 8(11):1015-1019. 346. Tschoepe D , Flomberg M , Roesen P, Gries F A . Reduced platelet thromboxane formation after long-term administration of a dihydropyridine calcium channel blocker: a prospective, double-blind, placebo-controlled study with nitrendipine in borderline hypertensive patients with IDDM-type diabetes mellitus. Diabetes Research 1992; 19(3):125-131.  272  347. V a n Bortel L M , Schiffers P M , B o h m R O , Mooij J M , Rahn K H , Struyker et al. The influence of chronic treatment with verapamil on plasma atrial natriuretic peptide levels in young and elderly hypertensive patients. European Journal of Clinical Pharmacology 1990; 39 Suppl 1:S39-S40. 348. V a n Merode T, V a n Bortel L M , Smeets F A , Mooij J M , B o h m R O , Rahn K H et al. Verapamil and nebivolol improve carotid artery distensibility in hypertensive patients. Journal of Hypertension - Supplement 1989; 7(6):S262-S263. 349. V a n Merode T, V a n Bortel L , Smeets F A , B o h m R, Mooij J, Rahn K H et al. The effect of verapamil on carotid artery distensibility and cross-sectional compliance in hypertensive patients. Journal of Cardiovascular Pharmacology 1990; 15(1):109-113. 350. Vanhees L , Fagard R, Lijnen P, Amery A . Effect of antihypertensive medication on endurance exercise capacity in hypertensive sportsmen. Journal of Hypertension 1991; 9(11):1063-1068. 351. Viskoper JR, Laszt A , Faraggi D . Twenty-four-hour blood pressure control with isradipine in mild essential hypertension. American Journal of Hypertension 1991; 4(2 Pt2):161S-162S. 352. Viskoper JR, Laszt A , Farragi D . The antihypertensive action of isradipine in mild essential hypertension. Journal of Cardiovascular Pharmacology 1991; 18 Suppl 3:S911. 353. Viskoper JR, Laszt A , Faraggi D , Yaskil E . Does isradipine exert a special earlymorning blood pressure-lowering effect? Journal of Cardiovascular Pharmacology 1992; 19 Suppl 3:S74-S75. 354. Watts R W , Wing L M . A placebo-controlled comparison of diltiazem and amlodipine monotherapy in essential hypertension using 24-h ambulator)' monitoring. Blood Pressure 1998; 7(l):25-30. 355. Weber M A , Cheung D G , Graettinger W F , Lipson J L . Characterization of antihypertensive therapy by whole-day blood pressure monitoring. J A M A 1988; 259(22):3281-3285. 356. Webster J, Robb OJ, Jeffers T A , Scott A K , Petrie J C , Towler H M . Once daily amlodipine in the treatment of mild to moderate hypertension. British Journal of Clinical Pharmacology 1987; 24(6):713-719. 357. Webster J, Robb OJ, Jeffers T A , Scott A K , Petrie J C . Once-daily amlodipine in the treatment of mild to moderate hypertension. Journal of Cardiovascular Pharmacology 1988; 12 Suppl 7:S72-S75. 358. Webster J, Witte K , Rawles J, Petrie J C , Jeffers T A . Evaluation of a long acting formulation of nicardipine in hypertension by clinic and home recorded blood  273  pressures and Doppler aortovelography. British Journal of Clinical Pharmacology 1989;27(5):563-568. 359. Webster J, Petrie J C , Jeffers T A , Roy-Chaudhury P, Crichton W , Witte K et al. Nicardipine sustained release in hypertension. British Journal of Clinical Pharmacology 1991;32(4):433-439. 360. Weir M R , Chrysant S G , McCarron D A , Canossa-Terris M , Cohen J D , Gunter et al. Influence of race and dietary salt on the antihypertensive efficacy of an angiotensinconverting enzyme inhibitor or a calcium channel antagonist in salt-sensitive hypertensives. Hypertension 1998; 31(5):1088-1096. 361. Chrysant S G , Weder A B , McCarron D A , Canossa-Terris M , Cohen J D , Gunter P A . Effects of isradipine or enalapril on blood pressure in salt-sensitive hypertensives during low and high dietary salt intake. A J H 2000; 13: 1180-1188. 362. Welzel D , Burger K J , Weidinger G . Calcium antagonists as first-line antihypertensive agents: a placebo-controlled, comparative trial of isradipine and nifedipine. Journal of Cardiovascular Pharmacology 1990; 15 Suppl 1:S70-S74. 363. White W B , Mehrotra D V , Black H R , Fakouhi T D . Effects of controlled-onset extended-release verapamil on nocturnal blood pressure (dippers versus nondippers). COER-Veraparml Study Group. American Journal of Cardiology 1997; 80(4):469-474. 364. Wilson J, Orchard M A , Spencer A A , Davies J A , Prentice CR. Anti-hypertensive drugs non-specifically reduce "spontaneous" activation of blood platelets. Thrombosis & Haemostasis 1989; 62(2):776-780. 365. Winer N , Thys-Jacobs S, Kumar R, Davidson W D , Grayson M , Harris C et al. Evaluation of isradipine ( P N 200-110) in mild to moderate hypertension. Clinical Pharmacology & Therapeutics 1987; 42(4):442-448. 366. Winer N , Kirkendall W M , Canosa F L , Lewin A J , McMahon F G , Carr A A et al. Placebo-controlled trial of once-a-day isradipine monotherapy in mild to moderately severe hypertension. Journal of Clinical Pharmacology 1990; 30(11):1006-1011. 367. Wing L M , Russell A E , Tonkin A L , Watts R W , Bune A J , West M J et al. Mono- and combination therapy with felodipine or enalapril in elderly patients with systolic hypertension. Blood Pressure 1994; 3(l-2):90-96. 368. Wing L M , Russell A E , Tonkin A L , Bune A J , West M J , Chalmers JP. Felodipine, metoprolol and their combination compared with placebo in isolated systolic hypertension in the elderly. Blood Pressure 1994; 3 (1-2): 82-89. 369. Wing L M , Arnolda L F , Harvey PJ, Upton J, Molloy D , Bune A J et al. Lacidipine, hydrochlorothiazide and their combination in systolic hypertension in the elderly. Journal of Hypertension 1997; 15(12 Pt 1):1503-1510.  274  370. Woehler T R , E f f J, Graney W , Heald D , Ziemniak J , Magner D . Multicenter evaluation of the efficacy and safety of sustained-release diltiazern hydrochloride for the treatment of hypertension. Clinical Therapeutics 1992; 14(2):148-157. 371. Wysocki M , Persson B , Bagge U , Andersson O K . Flow resistance and its components in hypertensive men treated with the calcium antagonist isradipine. European Journal of Clinical Pharmacology 1992; 43(5):463-468. 372. Wysocki M , Andersson O K , Persson B , Bagge U . Vasoconstriction during acute hypervolemic hemodilution in hypertensive patients is not prevented by calcium blockade. Angiology 1998; 49(l):41-48. 373. Yamakado T, Oonishi N , K o n d o S, Noziri A , Nakano T, Takezawa H . Effects of diltiazem on cardiovascular responses during exercise in systemic hypertension and comparison with propranolol. American Journal of Cardiology 1983; 52(8):1023-1027. 374. Yasky J, Godoy J, Soutric J L , Arenoso H J . [Ambulatory ergometric and pressurometric results with a daily dose of 240 mg of diltiazem in labile hypertensive patients with threshold ergometry]. [Spanish]. Medicina 1991; 51 (2):111-120. 375. Zito M , Abate G , Cervone C, Squassante L , Calabrese G . Effects of antihypertensive therapy with lacidipine on ambulatory blood pressure in the elderly. Journal of Hypertension - Supplement 1991; 9(3):S79-S83. 376. Duval S, Tweedie R. A nonparametric "trim and fill" method of accounting for publication bias in meta-analysis. Journal of the American Statistical Association 2000; 95(449):89-98. 377. Eichenwald K , Kolata G . A doctor's drug trials turn into fraud. The N e w Y o r k Times. [Onkne] 1999 May 17 [cited 2006 September 10]. Available from: U R L : http://query.nytimes.com/gst/fuUpage.html?sec=health&res=9E06ElD7143EF934A 25756C0A96F958260 378. Gueyffier F, Boutitie F, Boissel JP, Coope J, Cuder J, E k b o m T et al. I N D A N A : a meta-analysis on individual patient data in hypertension. Protocol and preliminary results. Therapie 1995; 50(4):353-62. 379. World Flealth Organization. International Clinial Trials Registry Platform (fCTRP). [Onkne] 2006 [cited 2006 September 12]. Available from: U R L : http://www.who.int/ictrp/en/http://www.who.int/ictrp/en/ 380. Pham B A ' , Piatt R, McAuley L , Klassen T P , Moher D . Is there a "best" way to detect and irunimize publication bias? A n empirical evaluation. Evaluation and die Health Professions 2001; 24(2): 109-125. 381. Song F, Eastwood A , Gilbody S, Duley L , Sutton A . Chapter 21: Publication and related biases. In: Stevens A , Abrams K , Brazier J, Fitzpatrick R and Lilford R,  275  editors. The Advanced Handbook of Methods in Evidence Based Healthcare. London: S A G E Publications; 2001. p. 371-385. 382.  Preston R A , Materson BJ, Reda D J , Williams D W . Placebo-associated blood pressure response and adverse effects in the treatment of hypertension. A r c h Intern M e d 2000; 160:1449-1454.  383.  Bekelman J E , L i Y , Gross C P . Scope and impact of financial conflicts of interest in biomedical research: a systematic review. J A M A 2003; 289(4): 454-465.  384. Chapleau M W . Chapter A35 - Arterial Barioreflexes, In: Izzo J L , Black H R , editors. Hypertension Primer, 3 Edition. Philidelphia: Lippincott Williams & Wilkins; 2003. p. 103-111. rd  385.  Jackson W F . Ion channels and vascular tone. Hypertension 2000; 35 (1 Pt. 2): 173178.  386. Bong A , Wong M M Y , Wright J M . Adverse cardiac effects of calcium channel blockers. The Cochrane Database of Systematic Reviews 2004, fssue 1. Art. No.: CD004527. D O L 10.1002/14651858.CD004527.  276  8. A P P E N D I C E S 8.1  A p p e n d i x A - T r i a l Selection F o r m  DATE OF REVIEW: CT  Reviewer (circle one):  MW  BH  DM  STUDY UNIQUE IDENTIFIER Other ID# PUBLICATION D A T E JOURNAL FIRST AUTHOR'S N A M E  INCLUSION  CRITERIA:  YES  NO  PATIENTS WITFI R E N A L FAILURE E X C U D E D  • • • • •  H Y P E R T E N S I V E PATIENTS (DBP>90 m m Hg, SBP>140 m m Hg)  i—i 1 1  CCBs AS M O N O T H E R A P Y  •  • • • • • • • •  RANDOMIZED DOUBLE-BLIND PARALLEL PLACEBO A R M >3 W E E K S D U R A T I O N  BP M E A S U R E D A T B A S E L I N E A N D B E T W E E N 312 W E E K S  I N C L U D E  •  1 1  E X C L U D E  (if "YES" to all above) Are authors to be contacted in order to decide inclusion/exclusion? Y /  UNCLEAR  • • • • • • • • • N  277  8.2  A p p e n d i x B - Standard data extraction form for each trial  DATA EXTRACTION FORM (use one form per trial)  Administration Details Paper title: Paper number: Study ID: Other references to which this trial may link with:  Extractor name: Characteristics of Included Studies F u n d i n g Source (Potential Bias) FOR  AGAINST  N O BIAS  UNCLEAR  Methods  Design of Study:  Method of randomization:  Concealment of randomisation: Was this concealment adequate/inadequate/unclear? Blinding:  Description of withdrawals or dropouts:  278  Jadad's score: Additional notes:  Participants Total eligible for inclusion into trial: Total number enrolled into trial: Number in treatment group(s): Number in placebo group:  ;  Numbers of withdrawals or dropouts (treatment/control): Numbers completing trial (treatment/control): Age (mean):  (range):  Sex: Ethnicity: Severity of hypertension (circle one): M i l d ( D B P 90-105 mmHg) Moderate ( D B P 105-115 mgHg) Severe ( D B P > 115 mmHg) Inclusion criteria: Diagnostic entry criteria:  SBP  •  DBP  Exclusion criteria:  279  Baseline characteristics:  Source of participants:  Co-morbid conditions: Rx:  Rx:  Placebo  Obesity Hyperlipidemia Diabetes Smoking Myocardial infarction Angina Stroke Sedentary lifestyle Left ventricular hypertrophy Additional notes:  Interventions Duration of placebo run-in period: Setting: Types:  •_ '  Duration of treatment: Compliance:  Measured? Y / N  % Patients compliant  How  compliant? Goal of therapy:  DBP  or  SBP  Additional notes:  280  Outcomes Outcomes:  Adverse events:  Additional notes:  Cross-over trials Run-in phase Treatment & duration  Control & duration  Washout phase. Treatment & duration  Control & duration  Additional notes:  Comparison(s) i n this trial:  281  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share