UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Blood pressure targets in the treatment of patients with elevated blood pressure Arguedas Quesada, José Agustín 2004

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

Item Metadata

Download

Media
831-ubc_2004-0354.pdf [ 4.73MB ]
Metadata
JSON: 831-1.0091643.json
JSON-LD: 831-1.0091643-ld.json
RDF/XML (Pretty): 831-1.0091643-rdf.xml
RDF/JSON: 831-1.0091643-rdf.json
Turtle: 831-1.0091643-turtle.txt
N-Triples: 831-1.0091643-rdf-ntriples.txt
Original Record: 831-1.0091643-source.json
Full Text
831-1.0091643-fulltext.txt
Citation
831-1.0091643.ris

Full Text

BLOOD PRESSURE TARGETS IN THE TREATMENT OF PATIENTS WITH ELEVATED BLOOD PRESSURE by Jose Agustin Arguedas Q. A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF T H E REQUIREMENTS FOR THE DEGREE OF MASTER IN SCIENCE in T H E FACULTY OF GRADUATE STUDIES (Department of Pharmacology and Therapeutics) We accept this thesis as conforming to the required standard Dr.$ M.J/right Associate Professor Dept of Pharmacology and Therapeutics Dr. ThormtfPerry Jr. Dr. Ken Bassett Clinical Assistant Professor Associate Professor Dept of Pharmacology and Therapeutics Dept. of Family Practice THE UNIVERSITY OF BRITISH COLUMBIA July 2004 © Jose A. Arguedas 2004 THE UNIVERSITY OF BRITISH COLUMBIA FACULTY OF G R A D U A T E STUDIES Library Authorization In presenting this thesis in partial fulfillment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Name of Author (please print) Zj Date (dd/mm/yyyy) Title Degree: M , S f Year: 'XsiQ^ Department of 9 \ s a v The University of British Columbia Vancouver, B C Canada O- IN f grad.ubc.ca/forms/?formlD=THS page 1 of 1 last updated: 20-M-04 A B S T R A C T Background In the management of patients with elevated blood pressure, the ideal blood pressure target to maximize the reduction in morbidity and mortality has not been established. The standard of clinical practice for many years has been a target of less than or equal to 90 mmHg for diastolic blood pressure. More recently the focus has been a target of less than or equal to 140 mmHg for systolic blood pressure. However, there has been a tendency during the last few years to recommend lower target blood pressures than those traditionally used. Objectives The specific aim of this systematic review was to identify all randomized controlled trials where participants were randomized to different BP targets and to determine if, in the treatment of patients with elevated blood pressure, "lower target" blood pressures (<135/85 mmHg) are associated with reduction in mortality and morbidity as compared with "traditional target" blood pressures (< 140-160 mmHg systolic and < 90-100 mmHg diastolic). Design Systematic review with meta-analysis. Search strategy Electronic search of MEDLINE (1966-2004), E M B A S E (1980-2004), and C E N T R A L (up to April 2004); references from review articles, clinical guidelines, and clinical trials. i i Selection criteria Randomized controlled trials in patients with elevated blood pressure randomized to "lower" or to "traditional" blood pressure targets and providing data on any of the primary outcomes. Analysis Two reviewers independently assessed and established the included trials. The primary outcomes were all-cause, cardiovascular and non-cardiovascular mortality; total serious adverse events; other cardiovascular serious adverse events; all other serious adverse events. The secondary outcomes were achieved mean systolic and diastolic blood pressure, percentage of patients achieving the target blood pressure levels, withdrawals due to adverse effects, and mean number of antihypertensive drugs per patient. Main results Six trials including 21,751 subjects were identified. Two trials included only patients with diabetes, and three trials included only patients with chronic renal disease. None of the trials compared different targets for systolic blood pressure. Therefore, at present we have no information regarding the benefits or harms of trying to achieve "lower targets" as compared with "traditional targets" for systolic blood pressure. In trials comparing diastolic blood pressure targets, despite a greater achieved reduction in blood pressure, trying to achieve the "lower targets" instead of the "traditional target" did not result in any change in total, cardiovascular or non-cardiovascular mortality, and did not result in any change in the incidence of myocardial infarction, stroke, congestive heart failure, the composite outcome of major cardiovascular events, or end-stage renal disease. The overall safety of the more intensive treatment cannot be assessed due to the ii i lack of information regarding total serious adverse events and withdrawals due to adverse effects. A sensitivity analysis performed in diabetic patients demonstrated that despite achieving significantly lower mean systolic and diastolic blood pressures, the groups allocated to a target diastolic blood pressure < 80 mmHg did not achieve a statistically significant benefit in any of the mortality and morbidity outcomes as compared with a target of < 90 mmHg. However, there was a trend toward decreased total mortality, major cardiovascular events and stroke incidence in diabetics randomized to "lower target" as compared with the "traditional target". A sensitivity analysis in patients with chronic renal disease demonstrated that despite achieving a substantially lower systolic and diastolic BP, there was no statistically significant difference in mortality, total cardiovascular events or end stage renal disease with "lower" as compared with "traditional" targets. Reviewers' conclusions In the absence of evidence, systolic blood pressure targets must be those that have been demonstrated to be better than placebo or no treatment in randomized controlled trials: < 150 to 160 mmHg. For the non-diabetic non-chronic renal disease patients with elevated blood pressure, the diastolic target should be < 90 mmHg. Treating patients to lower targets is not associated with a mortality or morbidity benefit. For the subgroups of patients with diabetes mellitus or chronic renal disease the conclusions are the same, but data are limited and the possibility exists that a clinically significant benefit or harm for lower targets could have been missed. Guidelines recommending lower systolic and diastolic blood pressure targets require testing in randomized trials, especially for patients with diabetes mellitus and chronic renal disease. iv T A B L E OF CONTENTS ' Abstract i i Table of contents v List of tables viii List of figures x Acknowledgments xi Dedication xii 1. INTRODUCTION 1 1.1 Background information 2 1.2 Objectives 16 1.2.1 Primary objective 16 1.2.2 Secondary objectives 16 2 M E T H O D O L O G Y 17 2.1 Design 18 2.2 Criteria for considering studies for this review 18 2.2.1 Types of studies 18 2.2.2 Types of participants 19 2.2.3 Types of interventions 19 2.2.4 Types of outcome measures 20 2.3 Search strategy for identification of studies 20 2.4 Methods of the review 21 2.5 Sensitivity analysis 22 v 3. RESULTS 23 3.1 Search results 24 3.1.1 Characteristics of included studies 25 3.1.1.1 Modification of Diet in Renal Disease (MDRD) 25 3.1.1.2 Totoetal 27 3.1.1.3 Hypertension Optimal Treatment (HOT) 29 3.1.1.4 Appropriate Blood Pressure Control in Diabetes (H) 31 3.1.1.5 Appropriate Blood Pressure Control in Diabetes (N) 33 3.1.1.6 African American Study of Kidney Disease and Hypertension trial (AASK) 35 3.1.2 Overall description of included studies 37 3.1.3 Methodological quality 38 3.1.4 Excluded studies 39 3.1.4.1 Treat Blood Pressure Better Study (BBB) 39 3.1.4.2 Hypertension in Diabetes IV Study (HDS) 40 3.1.4.3 United Kingdom Prospective Diabetes Study (UKPDS 38) 40 3.1.4.4 Lewis et al 41 3.2 Outcomes 43 3.2.1 Primary outcomes 43 3.2.1.1 Mortality 43 3.2.1.2 Total serious adverse events 46 3.2.1.3 Other cardiovascular serious adverse events 46 3.2.1.4 A l l other serious adverse events 49 3.2.2 Secondary outcomes 50 3.2.2.1 Systolic blood pressure achieved 50 3.2.2.2 Diastolic blood pressure achieved 52 3.2.2.3 Proportion of patients not achieving the blood pressure target 53 3.2.2.4 Withdrawals due to adverse effects 54 3.2.2.5 Number of antihypertensive drugs needed per patient 55 3.2.3 Summary 56 4. DISCUSSION 58 4.1 General discussion 59 4.2 Sensitivity analysis in patients with diabetes mellitus 64 4.3 Sensitivity analysis in patients with chronic renal disease 68 4.4 Limitations 70 4.5 Comparison with other published reports 72 4.5.1 Clinical guidelines 72 4.5.2 Blood Pressure Lowering Treatment Trialists' Collaboration 74 5. CONCLUSION 76 5.1 Final conclusions 77 5.1.1 Implications for clinical practice 77 5.1.2 Implications for research 78 6. BIBLIOGRAPHY 81 7. APPENDIX 1 90 vii LIST OF T A B L E S Table 1. Randomized controlled trials comparing drug therapy titrated to achieve a blood pressure target with placebo or no treatment in patients with mild to moderate elevations of blood pressure 5 Table 2. Goals of antihypertensive therapy according to several recently published clinical guidelines 13 Table 3. Summary of randomized controlled trials included 42 Table 4. Comparison of total, cardiovascular, and non-cardiovascular mortality between "lower" and "traditional" blood pressure targets 46 Table 5. Comparison of cardiovascular serious adverse events between "lower" and "traditional" blood pressure targets 50 Table 6. Summary of secondary outcomes 51 Table 7. Estimated percentage of patients by dose titration step according to target blood pressure group at 24 months of follow-up in the HOT trial 56 Table 8. Comparison of outcomes and achieved blood pressures in trials where the difference in the targets was at least 10 mmHg 62 Table 9. Outcomes in diabetic patients included in the HOT trial according to blood pressure target 64 viii Table 10. Distribution of patients with previous cardiovascular or cerebrovascular disease at baseline according to target blood pressure in the A B C D (H) and A B C D (N) trials 66 Table 11. Comparison of outcomes and achieved blood pressures between "lower" and "traditional" blood pressure targets in patients with diabetes mellitus 67 Table 12. Comparison of outcomes and achieved blood pressure between "lower" and "traditional" blood pressure targets in patients with chronic renal disease 68 ix LIST O F FIGURES Figure 1. Comparison of total mortality according to blood pressure targets 44 Figure 2. Comparison of cardiovascular mortality according to blood pressure targets. 45 Figure 3. Comparison of non-cardiovascular mortality according to blood pressure targets 45 Figure 4. Incidence of myocardial infarction according to blood pressure targets 47 Figure 5. Incidence of stroke according to blood pressure targets 47 Figure 6. Incidence of hospitalization due to congestive heart failure according to blood pressure targets 48 Figure 7. Incidence of end-stage renal disease according to blood pressure targets 48 Figure 8. Incidence of major cardiovascular events according to blood pressure targets 49 Figure 9. Achieved systolic blood pressure according to blood pressure target using the fixed effects model 52 Figure 10. Achieved diastolic blood pressure according to blood pressure target using the fixed effects model 53 Figure 11. Proportion of patients achieving the blood pressure target 54 Figure 12. Number of antihypertensive drugs needed per patient to achieve the target blood pressure 55 Figure 13. Mean decline in glomerular filtration rate (ml/min/1.73m2 per year) according to blood pressure target in patients with chronic renal disease 69 i x A C K N O W L E D G M E N T S I am very grateful to the following persons and institutions: My supervisor, Dr. James M . Wright, and the Thesis Committee members, Dr. Tom Perry Jr. and Dr. Ken Bassett, for their time and effort to teach me, but also for realizing that my family has needs beyond my academic interests. Universidad de Costa Rica, for the opportunity and the support to accomplish this valuable experience CONICIT, Ministerio de Ciencia y Tecnologia, Costa Rica, for the continuous support Dr. Marco Perez, independent reviewer in this project Dr. Fiona Turnbull, from the Blood Pressure Lowering Treatment Trialists' Collaboration, for providing useful additional information not available from the published reports of the trials Dr. Orlando Quesada Vargas, Facultad de Medicina, Universidad de Costa Rica, for his valuable guidance in my professional career Therapeutics Initiative, Department of Pharmacology & Therapeutics, The University of British Columbia, and especially to Mr. Ciprian Jauca and Mr. Stephen Adams, for their great help in many different ways Dr. Bernard McLeod, Professor, Department University of British Columbia, for his support M y family, for everything in my life of Pharmacology & Therapeutics, The xi DEDICATION A Tito y Tita, por su inmenso carino y dedication por ensefiarnos y motivarnos a navegar de la mejor manera posible en el rio de la vida. A Nena, Cfmcha y Dani, por ser mis fieles companeros en la aventura. Y a Viriam, Agus, Oqui y Andre, por ayudarme a guiar y a remar con alegria cuando pude tener mi propia canoa. 1. INTRODUCTION 1.1 BACKGROUND INFORMATION For many years elevated arterial blood pressure has been accepted as one of the major risk factors for cardiovascular disease [1-3]. Therefore, the primary goal of the management of patients with elevated blood pressure is to maximize the reduction in cardiovascular morbidity and mortality [4], and not just to reduce the numerical value of the blood pressure. Epidemiological studies have shown a continuous direct relationship between cardiovascular events and blood pressure [5]. The relationship has a greater slope with increasing levels of blood pressure. The lower threshold at which this relationship no longer applies has not been definitively identified [3]. Any numerical cut-off value above which elevated blood pressure (hypertension) is defined is arbitrary. The standard for diagnosis of arterial hypertension is based on consensus recommendations, which attempt to predict the blood pressure above which it is expected that treatment will provide more benefit than harm. The values were initially focused on diastolic blood pressure, but have added more emphasis on systolic blood pressure over the last decades. Due to its relatively brief and often serious clinical course, the benefits of drug therapy in patients with malignant hypertension and diastolic blood pressure above 140 mmHg were easy to establish initially. Demonstrating that drug therapy provided benefits in patients with milder degrees of elevated blood pressure have been more difficult. At the present time it remains unclear the threshold above which treatment benefits outweighs harm. When treating elevated blood pressure there are two critical and important decisions: 1) what is the threshold of blood pressure above which treatment is required? And 2) what target blood pressure is the goal of therapy? This second questions is the primary topic of this dissertation. It is the critical question that the practitioner in the trial or in clinical practice must answer to make treatment decisions. For example, i f the target is less than 2 or equal to 90 mmHg and the blood pressure is documented as 92 mmHg then the practitioner would increase the dose or add another antihypertensive drug. If the blood pressure is 90 mmHg then the practitioner would not change therapy. The standard of clinical practice for many years has been a target of less than or equal to 140 mmHg for systolic blood pressure, and a target of less than or equal to 90 mmHg for diastolic blood pressure. The first step in determining how this target came about is to look at the trials that have been used to provide the evidence that drug treatment for patients with elevated blood pressure is beneficial. The first major trial showing benefit occurred during the 1960's, when the Veterans Administration Cooperative Study (VA-I 1967) compared placebo with antihypertensive treatment using a diastolic blood pressure target lower than 90 mmHg [6]. This trial was conducted in 523 male patients with individual mean baseline diastolic blood pressure (DBP) from 90 through 129 mmHg. The study was terminated early, after 1.5 years, in the subgroup of 143 patients whose baseline mean diastolic blood pressure was 115 through 129 mmHg, because of the high incidence of cardiovascular events in the placebo group as compared with the treated patients. The remaining 380 randomized patients, whose averaged baseline DBP was between 90 and 114 mmHg, were followed for a longer time (VA-II 1970). In these patients it took almost 4 years to demonstrate a statistically significant difference between the groups [7]. Drug treatment with a DBP target lower than 90 mmHg as compared with placebo was associated with a decreased risk of developing a fatal or non-fatal cardiovascular morbid event. Placebo treated patients also progressed more often to higher levels of blood pressure. The degree of benefit was related to the level of pre-randomization blood pressure, with significantly larger absolute benefits for those patients with DBP between 105 and 114 mmHg than for those with DBP between 90 and 114 mmHg. A number of additional similar studies subsequently were conducted. Table 1 summarizes all trials where drug treatment was compared with placebo or no treatment 3 and for which the drug treatment group had a specific blood pressure target [8-20]. Inclusion criteria, blood pressure targets and results varied. Most trials focused on a diastolic blood pressure target. More recently, some trials have made similar comparisons in patients with isolated elevated systolic blood pressure, defined as a systolic blood pressure (SBP) greater than 160 mmHg, and a DBP lower than 90 or 95 mmHg. These trials had a systolic blood pressure target. In most of the published randomized controlled trials designed to test the benefits and harms of treatment titrated to achieve a specific blood pressure goal, the target was either a SBP below 150 mmHg or 160 mmHg for isolated systolic hypertension trials, or a DBP below 90 mmHg or 100 mmHg for most other trials (table 1). It is also important to appreciate that 30% to 40% of patients in these trials do not achieve the defined target, despite the requirement to titrate doses and add up to 3 or 4 antihypertensive drugs [21]. In spite of these facts, there has been a tendency during the last few years to recommend treatment of elevated blood pressure more aggressively toward much lower target blood pressures. This trend toward "the lower the pressure the better", as clearly expressed in an editorial accompanying the publication of the 2004 British Hypertension Society guidelines [22], assumes that treating to lower blood pressure targets will achieve an additional reduction in cardiovascular morbidity and mortality. 4 es cu S •s cu H—» o c o o •s o x -cu = cu I-I co CO CU I-I OH T 3 O CO > CU o cd T3 "5 CU 60 C "S OH O cu cd •c 3 co co CU I-i OH T3 O O O co Cl o > CU t3 CU 1 3 <U O a o o ts es o cu cu N a o CS cd es .a '-4-» OH 3 co CU t-i g "3 CO O *H—» CO •c CU H-» O cd , * '•S CU .g 13 co cd ICQ cu 60 cd" PQ es 5 .2 T3 OH 1 I OH cd CU ca H r-OS < > C! CU es cu > CU-I-I cu co cd > O & o o cu o cu J O 'o g i T3 CU CU -O T3 cu OH O O CO cd if o feb o -8 cu cd e! cu cd OH cd a CU H w tA 60 ^2 cd ^ cu CM CQ Q T3 U >. -*-» "cd o a I-I CU o cd OH 3 I CU •a cu 43 cu > ts es cu O 00 co o r - -OS > CU > +5 o CO CU60 es o cu >> • 1—1 cd a o 3 1 CO cd > O 5 cu CU cu 60 cd § cu CO I-I CU o T3 CU a T3 C3 cd cu <H-H I cu 60 cu cu .3 - 5 CU CO CO CU T 3 CU td es o 'co C3 CU tt CU cu I-I cu > cu CO OH o > cu T3 CU I-I co cd T3 O O cu g .£> o cu cu cu cu es o "is N a o C3 cd i n cu I-I OH <+H O rr a o as OH PQ Q t/i cd CO C3 O '3 o •T—I "E, a o o c+i O CU o s 'o _g cu H 5 cu o CN CN 00 ON 'H-I I o -8 cu "BH cu 2 o 1 es cu td OH ^1 O cd O as 60 o o Al OH PH Q 3 co CD g 2 CO CD *-4-» 'C U -4-» o CO , ^ CD OJ _c cd IPQ CD | P H ffl c o « H CM cd CD 05 CD a 1 3 o CD o 2 " 3 S3 o '% c o • • H o CD co nc 1 3 CD CO in ffe i 3 1 3 O a 00 Cd H 2 "o co I-I ca CD O CN c CD 60 a a oo V P H PQ Q CN CN © > CO •e o a <4H O CD -*-> cd l-i l-i CD O cd T 3 ca cd C H 1 3 CD -*-» cd CD ii H co CD O OO o r-o o\ CD U CD C CD m CD O 3 o CD I H CD O SS cd I a o CD 8 CO =3 O ' > CD t-i a, x CO I-l cd CD > > CD cd C H CD CD I i CD I i O •Sf 60 5 W A A P H CL, 99 m 00 co CD a o CD -*-» 3 O CD CD e CD o •a o cd y OQ o CD IH cd "3 X a, I CD ts CD CD _ > o cd CD CO CD O A CN o o i t H 13 1 CO cd > a -8 fe o cd CD1 3 <4H O CD CO O C H a o o CD Xi o oo ON 13 CD Id > J D 13 CD IH CD > CD co 1 3 9 c/f C3 _o cd o X CD ^ tH C H a O o o cd cd o 60 X o o CN V -o 9 co CD | H C H 1 3 O O 3 CD I-l c '3 CO §• I CD CD CD O C CD I i O fl CO C C CD ca o S3 Jfi 3 co cd > O T3 ^ cd o 1-3 ^ ca •c o 6 "3 -*-» o CD x £ .g1 3 O 6 ii ca 3 CO ca > o ca o ii o "cc? a G _o CD 3 1 3 CD I H c3 1 3 ca X o & CD -3 CD X CD > • t—i -*-» O ca CD o CD l i ca ^ CD 3 1 3 ca ca o e ca 1 3 s of CD co 3 CD S33 CO ca > o ca o T3 s "3 C»H I el o c C M O CD CD G CD 12 'o _C CD s .s G o CO * J ca »rt " O I-I CO c "o C M C O *-*-» O 3 1 3 CD I-l ca 1 3 ca X & 73 £ o to c CD 1 - 1 I H I H 3 3 o CO ca > o CD » O ca CD X H 1 _o '-J3 o 3 1 3 CD I H ca 1 3 ca X CH 3 O 1 3 CD •3 CD H CD _ > '-*-» CD ca CD X H 8 I 3 3 CO ca > o ca CD is ca .a 1 3 G ca CD o CO CD 03 •c CD -4-» CD ca CD C 03 ca ICQ CO I H ca CD > , O N I o G CD CD >» O N i o ro eo S3 -O N PH PQ Q 60 K a a o o CN V PH CQ on H H O CD CD G CD 1 3 • i—I > CD "3 CD CD O CD co ca CD CO 3 3 CO ca > o ca CD ca CD >^  o A co S3 ro 00 X O N o o O N PH PQ Q CD 60 S3 -t—• I PQ 60 X a © v PH PQ 00 60 X o O N V PH PQ Q G O I H , 3 Q 05 S3 CD IvO I in in o • 1—I •3 | -t—» a o U CD a, CU ca H i n oo o 00 O N O co o CN CO X o oo O N o oo PH PQ PH m t> oo § 2 in ro m oo O N O oi O o u L—1 3 CO CU P 'c3 CO O 'H—» CO •c 1) H - » o cd & H P o CO . p 13 CO cd co 60 +-» P H CQ p p o | 1 P H D H cd •c t—1 cd CO cd TD cd H P OH I TD CO td CL) I H CO > • *H H-J o cd CO & cd co >> •<* i oo ON C O cd P cd co o •5» c c o * H - » o 3 CO P H CQ 00 X v P H CQ Q 8 > CO 1 co cd > O 1 60 X B B OS o CN i o O CN S 2 cd T3 cd H P P-, o TD CO -t-t cd co t H co _> '-*-» O cd co H P H cd P H I H cS 60 B B o C O as as P H W 00 C4H o CO o p CO 12 ' o . p CO T P ^T! * * H P o 3 CO P 00 A P H CQ 00 CO p 13 co cd H O T3 PT o o cS P cd cd o o B co" o as 60 CN X T 3 P cd O o vp as V CO een BP Q ye. bel 60 X o P H CQ § A OO B H H O P CO ) H cd 60 X CO I H o •c P* CO > CO c2 60 X B B o CN -*-» CO cd CO P H CQ oo CO . P 13 CO cd H P r - -p cd O 6 -g CO ^ a td ^ CO CO O H P CO H P T3 cd H P o T3 CO td CO o cd co •co o \ p Al H P co co P H CQ 00 co o 3 TJ CO CN O S 3 W i -*-» c / l 00 H H co" CO H H O CO td I H CO - p o P H TD P CO o cd 1 o - C <H cd 3^ > c o * H - » o 3 TD CO O S CN TJ p cd O v© P -CO CO CO - P P H CQ 00 p o p TD P cd 60 X as V P H CQ Q 60 X 60 O >/-> as as Similarly, lower blood pressure targets than the standard have been set for specific groups of patients in several clinical guidelines. For example, it is widely proposed that more aggressive control of elevated blood pressure would produce additional benefits in diabetic patients. These guidelines are based on the hypothesis that this will lead to a reduction in the incidence of macro vascular and microvascular complications [23]. Both diabetes and elevated blood pressure are independent risk factors for the development of atherosclerotic disease. Scientific research performed over the last decade has proposed a large number of potential mechanisms for the molecular processes linking risk factors to the development of atherosclerotic plaques and atherothrombotic complications. These potential mechanisms include reduced endothelial-dependent vasodilation, down-regulation in the expression of nitric oxide synthase, increased release of proinflammatory cytokines and growth factors, increased interaction of platelets and monocytes with endothelial cells, insulin resistance, elevated plasma levels of antifibrinolytic substances such as PAI-1 (plasminogen activator inhibitor 1), increased tissue levels of angiotensin II, enhanced lipooxygenase activity, increased oxidative load, etc [24, 25]. Elevated blood pressure and diabetes are thought to be associated with several of these mechanisms [26]. These theoretical mechanisms plus epidemiological data showing a greater incidence of cardiovascular events in diabetic patients with elevated blood pressure compared with non-diabetic hypertensive patients have been used to justify lower targets for systolic and diastolic blood pressure in the treatment of diabetic patients. It has also been hypothesized that elevated blood pressure accelerates the loss of renal function in patients with chronic renal parenchymal disease. Therefore, it has been proposed that more intensive antihypertensive treatment aiming at lower blood pressure targets would result in additional cardiovascular and renal protection in those patients [27]. 9 Many clinical guidelines for the treatment of elevated blood pressure published during the last years have defined specific treatment goals based on blood pressure targets. The recommendations have not been homogeneous. The targets recommended in some of the more recently published guidelines are mentioned below and are summarized in table 2: 1- The Joint National Committee-7 Report (JNC 7 2003) recommends "treating systolic BP and diastolic BP to targets that are less than 140/90 mm Hg", but "because most patients with hypertension, especially those aged at least 50 years, will reach the diastolic BP goal once systolic BP is at goal, the primary focus should be on achieving the systolic blood BP goal" [28]. "In patients with hypertension with diabetes or renal disease, the BP goal is less than 130/80 mm Hg". 2- The 2003 European Society of Hypertension - European Society of Cardiology guidelines for the management of arterial hypertension (ESH-ESC 2003) recommends "that blood pressure, both systolic and diastolic, be intensively lowered at least below 140/90 mmHg and to definitely lower values, i f tolerated, in all hypertensive patients, and below 130/80 mmHg in diabetics, keeping in mind, however, that systolic values below 140 mmHg may be difficult to achieve, particularly in the elderly". Regarding patients with deranged renal function, it only refers to diabetic patients stating that "renal protection in diabetes has two main prerequisites: first, to attain very strict blood pressure control (<130/80 mmHg; and even lower, < 125/75 mmHg, when proteinuria > 1 g/day is present); and secondly, to lower proteinuria or albuminuria to values as near to normal as possible" [29]. 3- The 2003 World Health Organization (WHO) / International Society of Hypertension (ISH) (WHO/ISH 2003) states that "the primary goal of therapy is to lower systolic blood pressure, and the pragmatic target of below 140 mmHg is reaffirmed", and "the diastolic blood pressure to about 90 mmHg" [30]. There is also a small section entitled "Targets for blood pressure lowering in hypertensive patients at high risk" referring to patients with established cardiovascular disease, diabetes, and renal 10 insufficiency; for them the conclusion is "a target of <130/80 mmHg seems appropriate" [30]. 4- The 2004 American Diabetes Association (ADA 2003) standards of medical care states that "patients with diabetes should be treated to a systolic blood pressure <130 mmHg", and "to a diastolic blood pressure < 80 mmHg" [23]. 5- Finally, the British Hypertension Society guidelines for hypertension management 2004 (BHS 2004) states that "for most patients a target of < 140 mmHg systolic blood pressure and < 85 mmHg diastolic blood pressure is recommended. For patients with diabetes, renal impairment or established cardiovascular disease a lower target of < 130/80 mmHg is recommended" [31]. Attempting to achieve lower blood pressure targets has several consequences. The most obvious is the need for large doses of and increased number of antihypertensive drugs. This has costs to patients in terms of inconvenience and economic costs. It also will increase adverse drug effects, which i f serious could cancel any potential benefit associated with any lower blood pressures achieved In addition, there is the potential that lowering blood pressure too much can be related to an increase in cardiovascular events. Some observations have suggested that excessive lowering of the diastolic blood pressure with drugs can be associated with an increased number of deaths due to coronary heart disease [32]. This relationship, called the "J-curve phenomenon", was initially described in hypertensive patients with ischaemic heart disease, and the curve's inflection point was said to be around 85 mmHg [33]. More recently, an increased risk of stroke was described with treated diastolic blood pressure lower than 65 mmHg as compared with higher pressures [34]. Finally, a recent post-hoc subgroup analysis of the HOT trial has raised concern about the possibility that a J-shaped curve exists in smokers with elevated blood pressure due to a significantly increased risk of total and cardiovascular deaths, major cardiovascular events and stroke in these patients randomized to diastolic blood pressure < 80 mmHg or < 85 mmHg compared with < 90 mmHg [35-36]. 11 1W; 5 v ^ W ^W^ k 12 c CU CU co ca CJ co CCS el cu ca Q CL . M o oo © CO 60 o oo o co VI c cu ca O H CO a CO cu cu el .2 O 00 O CO o oo © CO V OX) X © 00 o CO VI ca >-cu el cu I O 60 X o O N O CO 00 g CU ca el o CO 3 O P H ca o 60 OH CQ 13 cu ts IH CU 60 X e © © T f co >> CO ca o 60 •c C H CU E—1 cu n CO CO CU VH C H T3 O O 3 60 1 oo © T f VI cu C 15 I c o 'co c cu C+H o -4-» CU 'o O oo cu C H O BJ co o o CN >> 60 O ' "o S3 U C+H O -*H CU 'o o 00 CU CH 2 w T f © © C N c o o o co co < CO CU -4-» CU o •g I T f o o CN cu • i—i o o 00 el o 'co c CU •e CU J3 CO %-» •c PQ 14 In summary, the ideal blood pressure target in the treatment of patients with elevated blood pressure has not been established. Furthermore, the assumption that lowering blood pressure by pharmacological means results in the same cardiovascular risk reduction as that associated with similarly lower blood pressures recorded in epidemiological studies needs to be demonstrated in randomized controlled trials. The only way to prove that a lower target is beneficial is a trial where patients are randomized to different treatment targets. Our goal therefore was to identify all randomized controlled trials where patients were randomized to different targets and to find out i f targets lower than traditional targets had any advantages. For purposes of this or any systematic review it is important to precisely define those categories. The "lower target" category is defined as any systolic blood pressure target less than or equal to 135 mmHg, and any diastolic blood pressure less than or equal to 85 mmHg. The "traditional target" category is defined as a systolic blood pressure target less than or equal to 140-160 mmHg, and a diastolic blood pressure target less than or equal to 90-100 mmHg. We have chosen a range for both target categories to be inclusive so as to not lose trials and because in the trials making up the evidentiary base different targets were used and the "traditional target" ranges incorporate most of them. Treatment targets higher than 160 mmHg systolic and higher than 100 mmHg diastolic are not eligible because they are considered to be inappropriately high. Finally, these categories are mutually exclusive, which means that targets included in the "lower target" category cannot overlap with the "traditional target" category. For example, targets of < 125 mmHg systolic or 75 mmHg diastolic are included in the "lower target" category, whereas targets of < 150 mmHg systolic or < 95 mmHg diastolic are included in the "traditional target" category. The specific aim of this systematic review is to identify all randomized controlled trials where participants were randomized to different BP targets and to determine if, in the treatment of patients with elevated blood pressure, "lower target" blood pressures, as 15 defined above, are associated with reduction in mortality and morbidity as compared with "traditional target" blood pressures as defined above. 1.2 OBJECTIVES 1.2.1 Primary objective "Lower targets" are defined as all targets less than or equal to 135 mmHg systolic and/or less than or equal to 85 mmHg diastolic (<135/85 mmHg). "Traditional targets" are defined as all targets less than or equal to the range of 140 to 160 mmHg systolic and/or all targets less than or equal to the range of 90 to 100 mmHg diastolic (< 140-160/90-100 mmHg). The main objective of this investigation is to determine i f there is a reduction in total mortality and morbidity associated with "lower targets" as compared with "traditional targets" in the management of patients with elevated arterial blood pressure. The morbidity outcomes are listed below. 1.2.2 Secondary objectives 1. To determine i f there is a change in mean achieved systolic and diastolic blood pressure associated with "lower targets" as compared with "traditional targets" in patients with elevated blood pressure. 2. To determine i f there is a change in withdrawals due to adverse events with "lower targets" as compared with "traditional targets", in patients with elevated blood pressure. 16 M E T H O D O L O G Y 2.1 DESIGN The design chosen is a systematic review with meta-analysis. This technique is based on a thorough review of the literature about a single topic, combining the results of several studies into a single outcome measure. Meta-analyses are intended to integrate existing information, looking for generalisability, consistency and inconsistency of scientific findings, and also to increase the statistical power of studies as well as the precision to estimate the treatment effects [37-40]. This systematic review and meta-analysis was carried-out following the guidelines developed by the Cochrane Collaboration [41]. 2.2 CRITERIA FOR CONSIDERING STUDIES FOR THIS REVIEW Studies will be considered for this review based on the following criteria: a. Types of studies b. Types of participants c. Types of interventions d. Types of outcomes measures 2.2.1 Types of studies Only randomized controlled clinical trials will be considered. Trials cannot be blinded as to blood pressure targets because the treating physicians must know the target to which each patient has been assigned in order to make the proper adjustment in the therapy to achieve the blood pressure goal. 18 A l l trials that reported any of the outcomes will be included. Trials will not be limited by any concomitant disease, other factor or baseline cardiovascular risk. There will be no language restriction. 2.2.2 Types of participants Participants must be adults, with elevated blood pressure documented in a standard way on at least 2 occasions, or adults already receiving treatment for elevated blood pressure. Although any numerical definition in this area is arbitrary, based on current standards of medical care elevated blood pressure was defined as baseline resting systolic blood pressure greater than 140 mmHg, and/or diastolic blood pressure greater than 90 mmHg. For the same reason, for patients with concomitant diabetes mellitus or chronic renal disease the criteria were systolic blood pressure greater than 130 mmHg, and/or diastolic blood pressure greater than 80 mmHg. 2.2.3 Types of interventions Trials were included i f individuals were randomized to a "lower" target systolic/diastolic blood pressure (< 135/85 mmHg) as compared with a "traditional" target blood pressure (< 140-160 mmHg systolic, or < 90-100 mmHg diastolic). Mean arterial blood pressure (MAP) has been defined as one third of systolic blood pressure plus two thirds of diastolic blood pressure. For that reason, in studies only reporting mean arterial blood pressure, a M A P lower than 107 mmHg, 102 mmHg and 97 mmHg were accepted as similar to systolic/diastolic blood pressures lower than 140/90 mmHg, 135/85 mmHg and 130/80 mmHg, respectively. 19 2.2.4 Types of outcomes measures Primary Outcomes: The hierarchy for primary outcomes will be: 1. All-cause mortality plus cardiovascular and non-cardiovascular mortality separately 2. Total serious adverse events 3. Cardiovascular serious adverse events, including myocardial infarction, stroke, congestive heart failure, chronic renal failure. 4. A l l other serious adverse events. Secondary Outcomes: The following are the secondary outcomes: 1. Systolic blood pressure achieved 2. Diastolic blood pressure achieved 3. Proportion of patients not achieving the target blood pressure levels 4. Withdrawals due to adverse effects 5. Number of antihypertensive drugs needed per patient. 2.3. SEARCH STRATEGY FOR IDENTIFICATION OF STUDIES The search was directed to randomized controlled trials comparing different blood pressure targets. Search included the following terms: randomized controlled trial, controlled clinical trial, random allocation, comparative studies, evaluation studies, follow-up studies, prospective studies, hypertension, arterial hypertension, high blood pressure, elevated blood pressure, hypertensive patients, target level, target blood 20 pressure, target systolic blood pressure, target diastolic blood pressure, intensive treatment, intensive blood pressure treatment, intensive control, intensive blood pressure control, tight control, tight blood pressure control, strict control, strict blood pressure control. A detailed description of the search is provided in appendix 1. The following databases were reviewed: MEDLINE from 1966 to April 2004 E M B A S E from 1980 to April 2004 C E N T R A L (Cochrane Central Register of Controlled Trials) up to April 2004 Reference lists existing in review articles, clinical guidelines, and clinical trials were also browsed for any study that may have not been identified by the search strategy. 2.4 METHODS OF THE REVIEW The outcomes to be compared and the trial eligibility criteria were specified before the result of any contributing trial was known. Two independent reviewers assessed and determined which trials were included or excluded. Discrepancies were resolved by discussion or by a third individual (JMW) i f necessary. Data from the trials was extracted independently by 2 reviewers from the included trials. For the synthesis and analysis of the data, Cochrane review manager software, RevMan 5.2, was used. Quantitative analyses of outcomes were based on intention-to-treat principle. Risk ratio (RR) and a fixed effects model were used to combine outcomes across trials. A standard chi-square statistic was used to test for heterogeneity of treatment effect between the trials [42]. A chi-square value < 0.05 was considered indicative of significant heterogeneity. A random effects model was used to test for statistical significance when significant heterogeneity existed [42-43]. In order to avoid 21 an inflated chance of a type I error due to the multiple comparisons performed, an alpha value of < 0.01 was chosen to combine outcomes across trials. The assessment of quality of the trials was based on considering the four potential sources of systematic bias in trials that need to be assessed for a review according to the Cochrane Reviewers' Handbook: selection bias, performance bias, attrition bias, and detection bias. 2.5 SENSITIVITY ANALYSES The following sensitivity analyses were performed: a. In order to maximize any potential difference between the "lower" and the "traditional" targets, a sensitivity analysis was performed comparing only those trials where the difference in the targets was at least 10 mmHg. b. Based on the previously mentioned epidemiological and physiopathological factors, it has been suggested that elevated blood pressure may have different implications in patients who also have concomitant diabetes mellitus or chronic renal disease. Therefore, sensitivity analyses were also performed for patients with these associated conditions. 22 J 3. RESULTS 23 3.1 SEARCH RESULTS The M E D L I N E search 1966-2004 identified 413 references. The E M B A S E search 1980-2004 identified 486 references. Searching the Cochrane Central Register of Controlled Trials database identified 137 references. Complete information of search results is provided in appendix 1. References regarding blood pressure targets provided in several recent review articles [44-53] and clinical guidelines [23, 28-31, 54, 55] were also retrieved. After reading the abstract or the complete report of these references, the majority of them were rejected; most of them were review articles or cohort studies. This left 20 references, which appeared to be appropriate for this systematic review. The detailed analysis of those 20 papers revealed: - 6 randomized controlled trials from 16 publications that met the inclusion criteria for this systematic review - 4 randomized controlled trials from 4 publications that did not meet the inclusion criteria The six included trials were: 1- Modification of Diet in Renal Disease trial (MDRD 1994) [56]. Two additional publications of the same trial were identified [57-58]. 2- Toto et al (Toto 1995) [59]. 3- Hypertension Optimal Treatment trial (HOT 1998) [60]. Seven additional publications of this trial were identified [35, 61-66]. 4- Appropriate Blood Pressure Control in Diabetes trial in hypertensive patients (ABCD-H 1998) [67], complemented by an additional publication [68]. 5- Appropriate Blood Pressure Control in Diabetes trial in normotensive patients (ABCD-N 2002) [69]. 6- African American Study of Kidney Disease and Hypertension trial ( A A S K 2002) [70]. 24 The main characteristics of those trials, chronologically ordered by date of first publication, are described in detail below and are summarized in Table 3. The reasons for excluding the four remaining randomized trials are also described below. 3.1.1 Characteristics of included studies 3.1.1.1 Modification of Diet in Renal Disease trial (MDRD 1994): a. Methods: Randomized, open label, controlled trial. Patients were randomized to two different blood pressure targets and also to two different protein diets. Blood pressure was measured monthly. The time of day when blood pressure was measured was not provided. The same nurse or technician took the sitting blood pressure three times in a quiet room, and the average of the last two was used as the blood pressure for the visit. The recommended antihypertensive regimen was an A C E inhibitor with or without a diuretic agent. A calcium channel blocker and other medications could be added as needed. The mean follow-up was 2.2 years. b. Participants: 840 patients with various chronic renal diseases were included. Chronic renal disease was established as a creatinine clearance of less than 70 ml per minute per 1.73 m of body surface area. To be included the participants had to be between 18 and 70 years 25 old, and have a mean arterial pressure of 125 mmHg or less. Mean arterial pressure was calculated as one third of systolic blood pressure plus two thirds of diastolic blood pressure. Arterial hypertension was not an inclusion criterion, but 86% of included participants were described as hypertensives; how hypertension was defined was not provided. The main exclusion criteria were pregnancy, insulin requiring diabetes mellitus, weight severely over or under normal, and urinary protein excretion exceeding 10 grams per day. c. Interventions: Patients were randomly assigned to a "usual"- or "low-blood pressure" group. "Usual blood pressure" was defined as a mean arterial pressure equal to or less than 107 mmHg (approximately < 140/90 mmHg) for patients < 60 years of age and < 113 mmHg (approximately < 160/90 mmHg) for patients > 60 years, whereas "low blood pressure" was defined as a mean arterial pressure lower than or equal to 92 mmHg (approximately < 125/75 mmHg) for patients < 60 years and < 98 mmHg (approximately < 135/80 mmHg) for > 60 years. d. Outcomes: The rate of change in glomerular filtration rate was the primary outcome measured. Other recorded outcomes were death, end-stage renal disease requiring dialysis or transplantation, and other serious medical conditions. e. Additional notes: Patients were also randomized to usual- or low-protein diet. 26 3.1.1.2 Toto et al (Toto 1995): a. Methods: The study was a 2 X 2 factorial, randomized controlled trial. Patients were randomized to either placebo or enalapril and to either "strict" or "conventional" blood pressure ranges. Before randomization, diastolic blood pressure was lowered to 80 mmHg or less over a 3 to 6 months initial assessment period. Patients able to achieve that target were randomized and included in the study. Blood pressure was measured in the supine position with a mercury sphygmomanometer after a minimum of 5 minutes rest. The time of day when blood pressure was measured was not provided. Three measurements were taken at 2-minute intervals. The mean of those measurements was used. Patients were randomly assigned to receive enalapril or placebo. In addition, to achieve the target diastolic blood pressure a stepped-care approach with antihypertensive medications was used: a diuretic was the initial drug, followed by a beta-blocker, hydralazine or minoxidil, and clonidine, alpha-methyldopa or an alpha-1 blocker. With the exception of the diuretic, the maximum dose of each agent was used before moving to the next step. In patients assigned to "conventional" group, diastolic blood pressure was allowed to increase to the 85 to 95 mmHg range, whereas it had to be maintained between 65 and 80 mmHg in those assigned to the "strict" group. Mean follow-up was 40.5 ± 1.8 months in the "strict" group, and 42.2 ± 2.1 months in the "conventional" group. b. Participants: 87 patients with hypertensive nephrosclerosis were initially considered for the trial. Their age ranged from 25 to 73 years. The inclusion criteria were a diastolic blood 27 pressure higher than or equal to 95 mmHg, a serum creatinine greater than 1.6 mg/dL but lower than 7.0 mg/dL and a glomerular filtration rate less than or equal to 70 ml/min/1.73m2, history of long-standing hypertension, an inactive urine sediment, a urinary protein excretion rate lower than 2 grams per day, and no physical or biochemical evidence for a humoral-mediated cause for hypertension. Exclusion criteria were diabetes mellitus, a recent history (in the previous 4 months) of malignant hypertension, stroke or myocardial infarction, acute renal failure of any cause, analgesic abuse, polycystic kidney disease and other causes of chronic renal disease, evidence of significant hepatic impairment, mental incapacity, pregnancy or lactation, primary hyperaldosteronism, renovascular hypertension, pheochromocytoma. Based on the initial assessment period, 77 patients were classified as "responders" and 10 patients were "non-responders". Since they were not randomized, "non-responder" patients were not included in this study. c. Interventions: "Responder" patients were randomized to either placebo or enalapril, in a double-blind design. They were also randomized to either "strict" or "conventional" blood pressure ranges in an open label design. "Strict" was defined as a diastolic blood pressure lower than 80 mm Hg, whereas "conventional" was defined as a diastolic pressure between 85 and 95 mm Hg. After randomization, the blinded study drug was titrated to maximum allowable dose and the unblinded antihypertensive agents were back-titrated as needed to achieve and maintain blood pressure control. d. Outcomes: The primary outcome was the rate of decline in glomerular filtration rate, measured by the renal clearance of 125I-iothalamate. Other outcomes were death, end-stage renal 28 disease and 50% decline in glomerular filtration rate or doubled serum creatinine (from baseline). e. Additional notes: Assignment to enalapril versus placebo did not change the results of the blood pressure control. The exclusion of patients not able to achieve the lower target during the pre-randomization period is a limitation of the trial as the results are only relevant to "responders" as defined in this study. 3.1.1.3 Hypertension Optimal Treatment trial (HOT 1998): a. Methods: Randomized, open label, controlled trial, with blinded endpoint evaluation (PROBE) design. A n Independent Clinical Event Committee, masked to the group allocation, evaluated all clinical events. Blood pressure was measured three times, by an oscillometric semiautomatic device, with the patient in the sitting position after 5 minutes of rest. The time of day when blood pressure was measured was not specified. Patients were randomly assigned to one of three diastolic blood pressure target groups. Block randomization was performed taking into consideration the following baseline variables: age, sex, previous antihypertensive therapy, smoking, previous myocardial infarction, previous coronary heart disease, previous stroke and diabetes mellitus. 29 A l l patients were treated with the same drugs in the same order. The following are the required steps allowed to attempt to achieve the target blood pressure: Step 1- felodipine 5 mg once a day Step 2- a starting dose of an angiotensin converting enzyme (ACE) inhibitor or beta-blocker is added Step 3- the dose of felodipine is increased to 10 mg once a day Step 4- the dose of the A C E inhibitor or the beta-blocker is doubled Step 5- a diuretic is added The average follow-up was 3.8 years. b. Participants: 19193 patients with elevated blood pressure, aged 50-80 years, were initially included, but the study population was composed of 18790 patients. Four hundred and three patients were excluded early in the trial because of the suspicion of incorrect inclusion. Baseline diastolic blood pressure between 100 mmHg and 115 mmHg on two occasions, at least one week apart, was an inclusion criterion. The main exclusion criteria were malignant hypertension, secondary hypertension, diastolic blood pressure > 115 mmHg, stroke or myocardial infarction within 12 months prior to randomization, decompensated congestive heart failure, other serious concomitant diseases which could affect survival during the next 2-3 years, patients who required a beta-blocker, A C E inhibitor or diuretic for reasons other than hypertension, patients who required antiplatelet or anticoagulant therapy, and insulin treated diabetics. 30 c. Interventions: Patients were randomly assigned to one of three diastolic blood pressure target groups: less than or equal to 90 mmHg, less than or equal to 85 mmHg, or less than or equal to 80 mmHg. d. Outcomes: The outcomes measured were: total and cardiovascular mortality, all (fatal and non-fatal) myocardial infarctions including silent infarctions, all (fatal and non-fatal) strokes, and major cardiovascular events (all myocardial infarctions plus all strokes plus other cardiovascular deaths). e. Additional notes: Patients were also randomly assigned to acetylsalicylic acid 75 mg daily or placebo. 24% of all investigators' reported events were rejected by the Clinical Event Committee. 3.1.1.4 Appropriate Blood Pressure Control in Diabetes trial H (ABCD-H 1998) a. Methods: Randomized, open label clinical trial. Patients were randomized to "intensive" versus "moderate" blood pressure control. They were also allocated to either nisoldipine or enalapril as the initial antihypertensive medication. If the target blood pressure was not achieved with increasing doses, then open-label antihypertensive medications were added in a step-wise fashion, initially with metoprolol, then hydrochlorothiazide or additional drugs, but neither a calcium channel blocker nor an A C E inhibitor. 31 Blood pressure recordings were obtained at the time peak drug levels were expected and were an average of three seated readings obtained at each visit. A n independent end point committee, which was blinded to the study intervention arms, reviewed all cardiovascular events. The follow-up period was 5 years. b. Participants: Four hundred and seventy patients, between the ages of 40 and 74 years, with type 2 diabetes mellitus and a diastolic blood pressure equal to or higher than 90 mm Ffg were included. Exclusion criteria included myocardial infarction or a cerebrovascular accident within the previous 6 months, coronary artery bypass surgery within the previous 3 months, unstable angina pectoris within the previous 6 months, congestive heart failure N Y H A class III or IV, a demonstrated absolute need for A C E inhibitors or C C B , and a serum creatinine level > 3 mg/dL. c. Interventions: Patients were randomized into two treatment arms consisting of "intensive" treatment with a diastolic blood pressure goal of 75 mmHg, and "moderate" treatment with a diastolic blood pressure goal of 80-89 mmHg. d. Outcomes: The primary end point was the change in 24-hour creatinine clearance. Secondary end points included cardiovascular events, retinopathy, clinical neuropathy, and urinary albumin excretion. 32 e. Additional notes: Patients were also randomized to either nisoldipine or enalapril as the initial antihypertensive medication. A test for interaction between study-drug assignment and blood-pressure-control strategy showed that no interaction was present. 3.1.1.5 Appropriate Blood Pressure Control in Diabetes trial N (ABCD-N 2002) a. Methods: Randomized, open label controlled clinical trial. Patients were randomized to "intensive" (10 mm Hg below the baseline diastolic blood pressure) versus "moderate" (80-89 mm Hg) diastolic blood pressure control. Patients in the "moderate" therapy group were given placebo, whereas patients randomized to "intensive" therapy received either nisoldipine or enalapril in a blinded manner as the initial antihypertensive medication. If the target blood pressure was not achieved with increasing doses, then open-label antihypertensive medications were added in a step-wise fashion, initially with metoprolol, then hydrochlorothiazide or additional drugs, but not a calcium channel blocker nor A C E inhibitor. Blood pressure recordings were obtained at the time peak drug levels were expected and were an average of three seated readings obtained at each visit. A n independent end point committee, which was blinded to the study intervention arms, reviewed all cardiovascular events. The follow-up period was 5 years. 33 b. Participants: Four hundred and eighty patients, between the ages of 40 and 74 years, with type 2 diabetes mellitus were included. A l l of them had a baseline diastolic blood pressure between 80 and 89 mmHg and were not receiving antihypertensive medications at the randomization visit. The main exclusion criteria were: myocardial infarction or cerebrovascular accident within the previous 6 months, coronary artery bypass surgery within the previous 3 months, unstable angina pectoris within the previous 6 months, congestive heart failure N Y H A class III or IV, a demonstrated absolute need for A C E inhibitors or CCB, and a serum creatinine level > 3 mg/dl. c. Interventions: Patients were randomized into two treatment arms consisting of "intensive" or "moderate" treatment. The goal in the "intensive" treatment group was to achieve a decrease of 10 mmHg below baseline in diastolic blood pressure (i.e. 70 to 79 mmHg), whereas the goal in the "moderate" treatment group was to maintain a diastolic blood pressure between 80 and 89 mmHg. d. Outcomes: The primary end point was the change in 24-hour creatinine clearance. Secondary end points included cardiovascular events, retinopathy, clinical neuropathy, and urinary albumin excretion. e. Additional notes: Patients randomized to intensive therapy received either nisoldipine or enalapril in a blinded manner as the initial antihypertensive medication. Patients in the moderate group 34 were given placebo. However, by the end of the study 117 patients (48%) initially randomized to moderate therapy required treatment (systolic blood pressure > 159 and/or diastolic blood pressure > 89 mmHg on two consecutive visits). These individuals were started on either nisoldipine or enalapril according to randomization at entry into the study with the goal of maintaining the systolic blood pressure < 160 mmHg and diastolic blood pressure < 90 mmHg. A test for interaction between study-drug assignment and blood-pressure-control strategy showed that no interaction was present 3.1.1.6 African American Study of Kidney Disease and Hypertension trial (AASK 2002) a. Methods Randomized 3 x 2 factorial trial. Participants were randomly assigned to 1 of 2 mean arterial pressure goals, and to initial treatment with a betarblocker, an A C E inhibitor or a dihydropyridine calcium channel blocker. Open-label agents were added sequentially to achieve the blood pressure goal. Three consecutive seated blood pressure readings were measured with a sphygmomanometer after at least 5 minutes rest, with the mean of the last 2 readings recorded. The time of day when blood pressure was measured was not reported. A l l cardiovascular events, including cardiovascular deaths and hospitalizations for myocardial infarctions, strokes, heart failure, revascularization procedures, and other hospitalized cardiovascular events were reviewed and classified by a blinded end points committee. The follow-up was 3 to 6.4 years. 35 b. Participants One thousand and ninety four participants (18 to 70 years old), self-identified as African-Americans, with diastolic blood pressure higher than 94 mmHg were included in the study. They also had to have a glomerular filtration rate between 20 and 65 ml/min per 1.73 m 2 , and no identified cause of renal disease other than elevated blood pressure. Exclusion criteria were known history of diabetes mellitus, urinary protein to creatinine ratio of more than 2.5, accelerated or malignant hypertension within 6 months, secondary hypertension, evidence of non-BP-related causes of chronic kidney disease, serious systemic disease, clinical congestive heart failure, or specific indication for or contraindication to a study drug procedure. c. Interventions Participants were randomized to a "usual" mean arterial pressure goal of 102 mmHg (approximately 135/85 mmHg) to 107 mmHg (approximately 140/90 mmHg) or to a "lower" mean arterial pressure goal' of less than or equal to 92 mmHg (approximately < 125/75 mmHg). d. Outcomes The primary analysis in the trial was based on the rate of change in glomerular filtration rate (GFR slope), assessed by renal clearance of I 125 iothalamate. The protocol also designated a main secondary composite outcome, which included any of the following: a confirmed reduction in GFR by 50% or by 25 mL/min per 1.72 m 2 from the mean of the two baseline GFRs; end stage renal disease (dialysis or transplantation); or death. 36 e. Additional notes Patients were also randomized to treatment with 1 of 3 antihypertensive drugs (metoprolol, ramipril, amlodipine). 3.1.2 Overall description of included studies Six randomized controlled trials were identified. Three of them were multi site studies (MDRD, HOT and A A S K ) . A total of 21751 subjects were included in the studies. They were ambulatory patients. Forty six percent of them were female. The sample-size-based weighted average age was 60.6 years; none of the trials included subjects older than 80 years. Five trials reported ethnicity (information not available from the HOT trial). White patients constituted 85% of subjects in M D R D , 25% in Toto, 83% in A B C D (H), and 90% in A B C D (N). Black subjects represented 8% in M D R D , 75% in Toto, 13% in A B C D (H), 7% in A B C D (N), and 100% in A A S K . The remaining subjects were described in the trials as "other groups". The prevalence of diabetes at baseline is known in every trial: 3% in M D R D , 8% in HOT, 100% in A B C D (H), 100% in A B C D (N), 0% in Toto and 0% in A A S K (the latter two trials excluded patients with diabetes). Other data on pre-existing conditions for subjects were not consistently reported in the studies. Four trials reported the baseline prevalence of coronary heart disease: 9.6% in M D R D , 7.5% in HOT, 25% in A B C D (H), and 14% in A B C D (N). The same four trials reported the baseline prevalence of previous stroke: 1.5% in M D R D , 1.2% in HOT, 2.5% in A B C D (H), and 3.5% in A B C D (N). Finally, chronic renal disease, defined as a measured glomerular filtration rate lower than 65 or 70 mL/min/1.73 m 2 , was present at baseline in 100% of patients from M D R D , Toto, and A A S K , but was not reported in the other three trials. 37 There was a large range in blood pressure values required for patients to be included in the trials. Four trials required elevated diastolic blood pressure, but the cut-off value varied: > 95 mmHg in Toto, > 100 mmHg but < 115 mmHg in HOT, > 90 mmHg in A B C D (H), and > 94 mmHg in A A S K . A B C D (N) only included patients with DBP between 80 and 89 mmHg. Finally, M D R D included patients with mean arterial pressure of 125 mmHg or less. For that reason, there was a great variability between trials in the mean baseline blood pressure: 170/105 mmHg in HOT trial, 156/98 mmHg in A B C D (H), 150/96 mmHg in A A S K , 136/84 mm Hg in A B C D (N), and 132/81 mmHg in M D R D trial. The trial by Toto reported only a baseline blood pressure equal to 123/76 mmHg recorded before randomization but after the initial assessment period performed to select "responders". The sample-size-based weighted baseline blood pressure was 166.4 mmHg for systolic and 102.9 mmHg for diastolic. A l l trials used a stepped care approach to achieve the assigned targets. There was a great variability in the type of drug used as initial therapy (see characteristics of included studies for specific details in each trial). The mean duration of follow-up ranged from 2.2 to 5 years. 3.1.3 Methodological quality Two of the multi-site studies (HOT and MDRD) reported that randomization was done at the study coordinating center. Blocked randomization was mentioned in M D R D , HOT, A B C D (H) and A B C D (N). Randomization was computer-generated in HOT, but the method of randomization was not described in the other 5 trials. 38 In Toto, the exclusion of patients not able to achieve the lower target during the pre-randomization period is a limitation of the trial as the results are only relevant to "responders" as defined in that study. None of the trials was blinded to blood pressure goal because of the need to titrate treatment to achieve the specific target. In HOT, 486 patients (2.6%) were lost to follow-up; they were equally distributed between the three target arms. The A A S K reported that 0 patients withdrew the study. Fourteen patients (1.6%) were lost to follow-up in the M D R D trial, but their distribution according to target is not provided. No specific information in this topic was provided in the remaining trials. HOT, A B C D (H), A B C D (N), and A A S K trials specifically stated that an independent clinical event committee, masked to the group allocation, evaluated all clinical events. Such information was not mentioned in the other two trials. 3.1.4 Excluded studies 3.1.4.1 Treat Blood Pressure Better Study or BBB for Behandla Blodtryck Battre in Swedish (BBB 1994) [71, 72] Randomized, open label, controlled trial. Two thousand one hundred and twenty seven hypertensive patients aged 45-67 years were included. To be included, they had to be receiving antihypertensive treatment, and their treated diastolic blood pressures on at least three consecutive visits were in the range between 90 and 100 mmHg. Patients were randomized to "intensified" or "unchanged" therapy. In the group allocated to "intensified" treatment, the purpose was to reduce the diastolic blood 39 pressure to < 80 mmHg. In the group allocated to "unchanged" therapy, the aim was to maintain the diastolic blood pressure in the range of 90-100 mmHg. This study, which showed no difference in morbidity or mortality outcomes between the target groups, was excluded from this meta-analysis because the number of patients randomized to each treatment arm was not provided and an attempt to contact the authors was unsuccessful. 3.1.4.2 Hypertension in Diabetes Study IV (HDS 1996) [73] Seven hundred and fifty eight hypertensive diabetic patients were included in this randomized controlled trial. The mean blood pressure at entry was 160/94 mmHg. This trial compared "tight control" of blood pressure aiming at a blood pressure of < 150/85 mmHg, with "less tight control" aiming at a blood pressure of < 180/105 mmHg. This study was excluded from the meta-analysis because the target for systolic blood pressure in the "tight control" group was higher than that stated in our protocol. In addition and more importantly the targets for both systolic and diastolic blood pressure in the "less tight control group" were much higher than specified in the protocol for this systematic review. These "less tight" pressures are similar to the escape criteria in most placebo or no treatment controlled antihypertensive trials, and much higher than conventional treatment goals prevalent since the 1970's. Furthermore, it is likely that participants in this trial represent a subgroup of patients included in UKPDS 38, because the study design is similar and the authors are the same. 3.1.4.3 United Kingdom Prospective Diabetes Study (UKPDS 38) [74] One thousand one hundred and eighty four hypertensive diabetic patients were included in this randomized controlled trial comparing "tight control" of blood pressure with "less tight control". The "tight control"group aimed at a blood pressure of < 150/85 mmHg. 40 In the "less tight control" group the target was originally set at < 200/105 mmHg, but was reduced to < 180/105 mmHg 5 years after the start of the study. This study was excluded from the meta-analysis because the target for systolic blood pressure in the "tight control" group was higher than that stated in our protocol. In addition and more importantly the targets for both systolic and diastolic blood pressure in the "less tight control group" were much higher than specified in the protocol for this systematic review. These "less tight" pressures are similar to the escape criteria in most placebo or no treatment controlled antihypertensive trials, and much higher than conventional treatment goals prevalent since the 1970's. 3.1.4.4 Lewis et al [75] Randomized controlled trial. 129 patients with type 1 diabetes mellitus and diabetic nephropathy were randomly assigned to a mean arterial blood pressure (MAP) goal < 92 mmHg or a M A P goal between 100 and 107 mmHg. The primary outcomes in this trial were surrogate markers of renal function in order to determine the impact of assignment to different levels of blood pressure control on the course of type 1 diabetic nephropathy. It was excluded because it did not provide data on any of the outcomes defined for this systematic review. The only reported clinical event was end-stage renal disease (ESRD). Twelve patients reached ESRD, but the distribution of those patients according to the blood pressure target assigned was not provided. It also reported achieved blood pressure but as mean arterial pressure, not as systolic and/or diastolic blood pressure achieved. The authors did not respond to written requests for that information. 41 CO el o i'S CH 6 \d IH O "SB X o ON © T f V J , X a s r-o VI "cd o 6 0 * C H < CO > 6 0 X a a ' in o\ I in oo S> P H pq Q co co > > 6 0 6 0 ffi X a a a a in oo VI VI O ON cd O 6 0 P H PQ Q ca o 6 0 P H Q 6 0 X m r--o 13 o 6 0 CN T f P H P H P H pq pq pq Q Q Q 1 3 cu 13 13 el cd s 1 3 OJ a o o 1 3 CD N O -a c CO IH WH O ccj 3 0 0 cn CU I o l l cd cu CO -4-» 'o • 1—1 P H CN CN cd (3 cu IH O S3 cu '-*-» cd C H o T f 00 CU co cd cu CO T f I cn cu .> 'co C cu •c cu 1^ cd -4H CU O o H cu CO cd cu co CI .2 o "cd '3 C H O o c3 in O N ON 00 ro ^ "cd S3 cu CM CU _> 'co e CU •e CU H C o ON r -oo 00 ON ON E - H O X in £3 cu CM O %-» CU jg -3 CU _> '55 C cu •c cu J 3 O r -T f oo ON ON Q O PQ < in cu cd CM CU H cu ,> 'co c cu o S3 o 00 T f CN o o CN Q O pq < T f NO o - M cn S3 cu cu •8 S3 O (3 T f ON O 0 0 cd CI CU cd o CM -g •2 P CU CU .> 'co s cu — i J 3 CN O o CN CU u co co cu IH CM o 13 o co cd P H CQ Q I co co CU IH CH "cd 'C CU S3 cd cu P H 3.2 OUTCOMES The results are discussed according to the hierarchy of outcomes previously mentioned. Several outcomes were not reported in the published trials. The missing information was requested by e-mail to the main authors of each trial, but it was not obtained. The authors of the A A S K trial replied that they are going to include the missing information in a paper they are preparing to submit for publication soon. There was no reply from the authors of the remaining trials. Some additional information, not included in the original published reports, was provided by the Blood Pressure Lowering Treatment Trialists' Collaboration [76]. That information is described for each outcome. 3.2.1 Primary outcomes 3.2.1.1 Mortality a. Total mortality Information from 5 trials was available for this analysis. The mortality data from the Modification of Diet in Renal Disease Study (MDRD) could not be included for the following reasons. The original publication of the M D R D trial mentioned 30 deaths, without specifying to which group they belonged [56]. A n additional publication of the same trial [58] provided the distribution, according to blood pressure goal, of 16 patients who died before reaching a study point, but not for 14 patients who died after a stop point but before their scheduled close-out visit The data on mortality from the A B C D (H) trial deserve an explanation. The first publication of the trial [67] mentioned 30 deaths in total, but it did not provide details on 43 the distribution according to blood pressure targets, because it focused on the comparison between the antihypertensive drugs used. A later publication stated that "patients randomized to intensive therapy had a lower incidence of all-cause mortality when compared to moderate therapy, 5.5% vs 10.7%, p= 0.037" [68], without providing absolute numbers. Given that 237 patients were assigned to the intensive treatment group, and 233 patients to the moderate treatment group, the absolute numbers of deaths calculated from the reported percentages would be 13 and 25 respectively, for a total of 38 deaths, which differs from the total mortality mentioned in the first report. Finally, the Blood Pressure Lowering Treatment Trialists' Collaboration (BPLTTC) reported 32 deaths in the same trial, 10 in the intensive treatment group and 22 in the moderate treatment group [76]. Due to the lack of concordance, the information from BPLTTC was used for this analysis, because it was the only one providing absolute numbers and because they were closer to the figures mentioned in the original report. As shown in Figure 1 and in Table 4, there was no difference in total mortality between the two blood pressure target groups: p= 1. The isolated decrease in total mortality found in the A B C D (H) trial was not repeated in the other trials. Figure 1. Comparison of total mortality according to blood pressure targets Study Lower target Traditional target nIH nIH RR (95%CI Fixed) Weight RR % (95%CI Fixed) AASK Trial ABCD (H) ABCD (N) HOT Toto 37/540 10/237 18/237 401 /12526 1 /42 Total(99%CI) 467/13582 Test for heterogeneity chi-square=6.01 df=4 p=0.2 Test for overall effect z=-0.05 p=1 43/554 22/233 20/243 188/6264 0/35 273 / 7329 12.6 6.6 5.9 74.7 -» 0.2 100.0 0.88[0.58,1.35] 0.45[0.22,0.92) 0.92(0.50,1.70) 1.07(0.90,1.27) 2.51(0.11,59.79] 1.00(0.82,1.21] .1 .2 1 5 10 Lower target Traditional target 44 b. Cardiovascular mortality The comparison for cardiovascular mortality does not include the trial by Toto et al, since cause of death was not given for the single patient who died in that trial. The information from the A B C D (H) trial was provided by the BPLTTC. There was no difference in cardiovascular mortality between the two blood pressure target groups: p= 0.8 (Figure 2, Table 4). Figure 2. Comparison of cardiovascular mortality according to blood pressure targets Study Lower target Traditional target nIH n/H RR (S5UCI Fixed) Weight RR (95%CI Fixed) AASK Trial ABCD (H) ABCD (N) HOT 14/540 6/237 13/237 186/12526 Total(99%CI) 219/13540 Test for heterogeneity chi-square=2.67 df=3 p=0.4S Test for overall effect z=0.32 p=0.8 16/554 11 /233 9/243 87/6264 123/7294 10.4 7.3 5.9 76.4 100.0 0.90[0.44,1.82] 0.54(0.20,1.43] 1.48(0.65,3.40] 1.07(0.83,1.38] 1.04(0.77,1.39] .1 .2 Lower target 5 10 Traditional target c. Non-cardiovascular mortality There was no difference in non-cardiovascular mortality: p= 0.7 (Figure 3, Table 4). The comparison does not include the trial by Toto et al for the previously specified reason. Figure 3. Comparison of non-cardiovascular mortality according to blood pressure targets Study Lower target Traditional target RR (9S%CI Fixed) Weight RR % (95%CI Fixed) AASK Trial ABCD (H) ABCD (N) 23/540 4/237 5/237 27/554 11/233 11/243 14.5 6.1 5.9 0.87(0.51,1.50] 0.36(0.12,1.11] 0.47(0.16,1.32] HOT 215/12526 101 Z6264 . i I- 73.5 1.06(0.84,1.35] Total(99%CI) 247/13540 Test for heterogeneity chi-square=5.65 df=3 p=0.13 Test for overall effect z=-0.40 p=0.7 150/7294 , • 100.0 0.96(0.73,1.26] .1 .2 Lower target 5 10 Traditional target 45 Table 4. Comparison of total, cardiovascular and non-cardiovascular mortality between lower and traditional blood pressure targets Outcome Number of Number of RR P studies participants (99% CI) Total mortality 5 20911 1.00 (0.82, 1.21) 1 C V mortality 4 20834 1.04 (0.77, 1.39) 0.8 Non-CV mortality 4 20834 0.96 (0.73, 1.26) 0.7 3.2.1.2 Total serious adverse events Total serious adverse events were not reported in any of the trials. 3.2.1.3 Other cardiovascular serious adverse events Each trial reported separately several cardiovascular serious adverse events. However, none of the events was reported in every trial. Fatal and non-fatal events were reported combined. A summary of reported cardiovascular serious adverse events is provided in Table 5. a. Myocardial infarction Only 3 trials reported the rates for myocardial infarction. To maintain homogeneity with the other trials for the purpose of this analysis, data from HOT do not include silent myocardial infarctions. As shown in Figure 4 and in Table 5, there was no difference in the risk of myocardial infarction: p= 0.15. 46 Figure 4. Incidence of myocardial infarction according to blood pressure targets Lower target Traditional target RR Weight RR Study n/H n/H (95%CI Fixed) % (95%CI Fixed) ABCD (H) ABCD (N) HOT 16/237 19/237 125/12526 14/233 15/243 84/6264 JBSL 10.0 10.5 79.5 1.12(0.56,2.25] 1.30(0.68,2.50] 0.74(0.57,0.98] Total(99%CI) 160/13000 Test for heterogeneity chi-square=3.13 df=2 p=0.21 Test for overall effect z=-1.44 p=0.15 113/6740 . —1 100.0 , 0.84(0.62,1.15] -1 1 1 1 1-.1 .2 1 5 10 Lower target Traditional target The incidence of silent myocardial infarctions reported in the HOT trial was not different between the lower (89/12526) and the traditional target groups (43/6264): RR 1.04, 99% CI (0.64, 1.67), p= 0.9. However, the fact that the trend is in the opposite direction as that seen with the clinically apparent myocardial infarctions is of potential significance. b. Stroke The information on stroke is available from only 3 trials. Data from A B C D (H) was provided by the BPLTTC. A lower incidence of stroke was reported in A B C D (N), but not in the other trials. There was no difference in the cumulative analysis for this outcome: p= 1 (Figure 5, Table 5). Figure 5. Incidence of stroke according to blood pressure targets Study Lower target n/H Traditional target n/H RR (95%CI Fixed) Weight % RR (95%CI Fixed) ABCD (H) ABCD (N) HOT 9/237 4/237 200/12526 9/233 13 /°13 6.2 8.7 85.1 0.98(0.40,2.43] 0.32(0.10,0.95] 1.06(0.83,1.36] 94/6264 I I rota!(99%CI) 213/13000 116/6740 Test for heterogeneity chi-square=4.44 df=2 p=0.11 Test for overall effect z=-0.05 p=1 I r r 100.0 0.99(0.74,1.34] -1 1 1 1 1-.1 .2 1 S 10 bjuisr target Traditional target 47 c. Congestive heart failure The only trials reporting data on congestive heart failure were the A B C D (H) and A B C D (N) trials. These were defined as congestive heart failure requiring hospital admission. Data from A B C D (H) was provided by the BPLTTC. There was no difference in this outcome: p= 0.9 (Figure 6, Table 5). Figure 6. Incidence of hospitalization due to congestive heart failure according to blood pressure targets Study Lower target Traditional target n/H n/N RR (95%CI Fixed) Weight % RR <95%CI Fixed) ABCD (H) ABCD (N) 9/237 12/237 Total(99%CI) 21 /474 Test for heterogeneity chi-square=0.04 df=1 p=0.83 Test for overall effect z=0.18 p=0.9 9/233 11 /243 20/476 45.5 54.5 100.0 0.98(0.40,2.43] 1.12(0.50,2.49] 1.06(0.48,2.32] .1 .2 Lower target 5 10 Traditional target d. End-stage renal disease End-stage renal disease, defined as the requirement of dialysis or kidney transplantation, was reported in the A A S K trial and in the trial by Toto et al. One hundred and six additional patients included in the M D R D trial developed end-stage renal disease, but they could not be included in this analysis because their distribution between "usual" or "lower" blood pressure was not provided. As shown in Figure 7, the incidence of end-stage renal disease was similar between patients randomized to "lower" or "traditional" blood pressure target: p= 0.8. Figure 7. Incidence of end-stage renal disease according to blood pressure targets Study Lower target Traditional target n/N nIH RR (95%CI Fixed) Weight % RR (95%CI Fixed) AASK Trial Toto 81 /540 7/42 Total(99%CI) 88/582 Test for heterogeneity chi-square=2.18 df=1 p=0.14 Test for overall effect z=-0.21 p=0.8 90/554 2/35 92/589 97.6 2.4 100.0 0.92(0.70,1.22] 2.92(0.65,13.15] 0.97(0.68,1.38] .1 .2 Lower target 5 10 Traditional target 48 e. Major cardiovascular events Information regarding major cardiovascular events, a composite outcome including myocardial infarction, stroke, heart failure or any cardiovascular death, is commonly reported in clinical trials. The BPLTTC's authors provided additional information for this outcome in the A A S K , A B C D (H) and A B C D (N) trials. The HOT trial reported major cardiovascular events, but did not include hospital admissions due to heart failure. Despite this, it was included in this analysis. The effect size does not change whether or not the HOT trial is included in the analysis. As shown in Figure 8 and in Table 5, none of the trials showed a difference in major cardiovascular events between patients randomized to the lower or the traditional target blood pressure: p= 0.5 for the comparison. Figure 8. Incidence of major cardiovascular events according to blood pressure targets Study Lower target nIH Traditional target n/N RR <95%CI Fixed) Weight % RR (95%CI Fixed) AASK Trial ABCD (H) ABCD (N) HOT 51 /540 36/237 37/237 451 /12526 60/554 39/233 39/243 232/6264 1 1 13.3 8.8 8.6 69.3 0.87[0.61,1.24] 0.91 [0.60,1.37] 0.97(0.64,1.47] 0.97(0.83,1.14] Total(99%CI) 575/13540 370/7294 •4 100.0 0.95(0.81,1.13] Test for heterogeneity chi-square=0.37 df=3 p=0.95 Test for overall effect z=-0.73 p=0.5 .1 .2 1 5 10 Lower target Traditional target 3.2.1.4 A l l other serious adverse events No other serious adverse events were reported. 49 Table 5. Comparison of cardiovascular serious adverse events between "lower" and "traditional" blood pressure targets Outcome Number Number of RR P of studies participants (99% CI) Myocardial infarction 3 19740 0.84 (0.62, 1.15) 0.15 Stroke 3 19740 0.99 (0.74, 1.34) 1 Congestive heart failure 2 950 1.06 (0.48,2.32) 0.9 End-stage renal disease 2 1171 0.97 (0.68, 1.38) 0.8 Major C V events 4 20834 0.95 (0.81, 1.13) 0.5 3.2.2 Secondary outcomes Some secondary outcomes were not available in several trials. Each of the outcomes is briefly described below. A summary is provided in Table 6. 3.2.2.1 Systolic blood pressure achieved Five trials reported the mean systolic blood pressure achieved (Figure 9). The M D R D trial was not included in the analysis because it reported mean arterial pressure, but not systolic blood pressure achieved. In HOT, achieved blood pressure was defined as the mean of all blood pressures from 6 months of follow-up to end of study. In A B C D (N) and in A B C D (H), the achieved blood pressure was defined as the average blood pressure for the last four years of follow-up. In A A S K it included blood pressure measurements after 3 months of follow-up. Toto et al. did not describe the way used to estimate achieved blood pressure. The standard deviation was not provided in the A B C D (H) trial. For that reason, it was imputed as the weighted mean from the remaining trials. 50 ro o o o o o o o V o o o o o OS oo ro oo ro ro O s r -ro oo CN CN ha o : l £ 1 03 CO 2 60 o CN co oo ID =5 Pi 03 CO XO CN IX) 03 C X IS '-1 l l 0) a o o H-» 03 C X 1 3 3 OS O CN 8 3 co CO K ex '"8 o |3 o CO CO > CO i a o ON o CN 3 CO CO e ex T 3 o o H O CJ CO > CO o < T i -ro Os CN 60 co S3 CO -c 0) ca o I i co 1 I CO ex IH CO ex CO u CO a '•8 CO CO X • CO o a e 1*3 CO s I-a CO i f . 53 Q The best estimate of the weighted mean change is obtained with the fixed effects model. Using this model, the weighted mean systolic blood pressure was 3.89 mmHg lower in the lower target group (Figure 9). Heterogeneity between trials was high. However, using the random effects model still achieved statistical significance with a lower systolic blood pressure in the lower target group: p= 0.003. Figure 9. Achieved systolic blood pressure according to blood pressure target using the fixed effects model Study Lower target TradKional target n rnean(sd) n mean(sd) WMD (95%CI Fixed) Weight WMD % <95%CI Fixed) AASK Trial ABCD (H) ABCD (N) HOT Toto 540 128 00(12.00) 237 132.00(11.75) 237 128.00(12.31) 12526 140.55(11.70) 42 133.00(19.40) 554 141.00(12.00) 4 233 138.00(11.34) 243 137.00(10.91) <-6264 143.70(11.30) 35 138.00(11.80) <-Total(99%CI) 13582 7329 Test for heterogeneity chi-square=202.23 df=4 p«0.00001 Test for overall effect 1=23.26 p«0.00001 5.3 2.5 2.5 89.5 0.2 100.0 -13.00[-14.42,-11.58] -6.00[-8.09,-3.91] -9.001-11.08,-6.92] -3.151-3.50,-2.80] -5.001-12.05,2.05] -3.891-4.33,-3.46] -10 -5 Lower target 5 10 Traditional target 3.2.2.2 Diastolic blood pressure achieved Five trials reported the mean diastolic blood pressure achieved, as defined above (Figure 10). In all of them, it was significantly lower in the group assigned to the lower target. The M D R D trial was not included because it reported mean arterial pressure achieved, but not diastolic blood pressure. The standard deviation was not provided in the A B C D (H) trial. For that reason, it was imputed as the weighted mean from the remaining trials. Using the fixed effects model as the best estimate of the weighted mean change, the weighted mean diastolic blood pressure was 3.38 mmHg lower in the lower target group (Figure 10). The inter-studies heterogeneity was high; therefore, the random effects 52 model was used to test for statistical significance. The weighted mean diastolic blood pressure was significantly lower in the lower target group: p< 0.00001. Figure 10. Achieved diastolic blood pressure according to blood pressure target using the fixed effects model Lower target TradKional target WMD Weight WMD Study n mean(sd) n mean(sd) (95%CI Fixed) % (95%CI Fixed) AASK Trial 540 78.00(8.00) 554 85.00(7.00) — 2.7 -7.00[-7.89,-6.11] ABCD (H) 237 78.00(5.17) 233 86.00(5.24) — 2.5 -8.001-8.94,-7.06] ABCD (N) 237 75.00(4.61) 243 81.00(4.68) — 3.1 -6.001-6.83,-5.17] HOT • 12526 82.15(5.05) 6264 85.20(5.10) • 91.4 -3.05[-3.20 ,-2.90] Toto 42 81.00(6.48) 35 87.00(5.92) 0.3 -6.001-8.77,-3.23] Total(99%CI) 13582 7329 • 100.0 -3.381-3.57,-3.19] Test for heterogeneity chi-square=215.07 df=4 p<0.00001 Test for overall effect 2=44.95 p<0.00001 -10 -5 0 5 10 Lou*r target Traditional target 3.2.2.3 Proportion of patients not achieving the blood pressure targets Only three trials, the A A S K , the HOT and the M D R D trials, reported the number of patients who did not achieve the target blood pressure levels. Data from the HOT trial was obtained from an interim analysis reporting results on blood pressure at 24 months of follow-up [64]. There was high heterogeneity between the trials, basically due to the results in A A S K . That difference can be explained because the definition for the traditional target in A A S K was a range of values rather than as any value below the target. This increases the possibility for missing the traditional target. Therefore, due to the lack of comparability, A A S K was excluded from this analysis. There was a highly significant larger proportion of patients not achieving the lower blood pressure target when HOT and M D R D were combined (figure 11): RR 2.36 (99% CI 2.17, 2.56), p<0.00001. 53 Figure 11. Proportion of patients not achieving the blood pressure target Lower target Traditional target RR Weight RR Study nIH n/H (95%CI Fixed) % (95%CI Fixed) HOT 4259/12526 940 / 6264 MDRD 168/432 20/408 Total(99%CI) 4427/12958 960 / 6672 Test for heterogeneity chi-square=30.25 df=1 p<0.00001 Test for overall effect z=26.72 p<0.00001 ^ 1.6 4 100.0 2.27[2.13,2.42] 7.93[5.09,12.36] 2.36[2.17,2.56] .1 .2 5 10 Lower better Traditional better 3.2.2.4 Withdrawals due to adverse effects None of the trials reported the total number of withdrawals specifically due to the treatment adverse effects in each treatment arm. Only fragmented information regarding adverse effects was provided in some of the trials. For example, in the M D R D trial, in 14 (3.2%) of patients in the lower blood pressure group and in 3 (0.7%) of patients in the higher blood pressure group, a reduction in antihypertensive medications was required due to persistent symptoms of hypotension (p= 0.01 for the comparison between groups) [58]. Furthermore, the blood pressure had to be raised because of persistent symptoms of hypotension in 17 patients, but their distribution according to blood pressure target was not provided [56]. However, the authors mentioned that no patient reached a stopping point in the study because of complications of hypotension [56]. The trial by Toto et al did not report information on this issue. The main report of the HOT trial did not provide information regarding withdrawals due to adverse effects. A preliminary publication of the same trial [64] described tolerability at 24 months and reported that there was no difference in overall incidence of side effects between the three target groups. However, according to the protocol, treatment was titrated in up to five steps, and the already cited preliminary publication mentioned that "more patients needed step 5 medication in the < 80 mmHg target group than in the < 90 54 mmHg target group", and "with increasing dose titration steps there was a gradual increase in the total number of side effects" [64]. In the A B C D (H) trial, reasons for discontinuation of the study were provided for the comparison between nisoldipine and enalapril, but not for the comparison between targets. No information was provided in the A B C D (N) trial report. The A A S K trial report did not include information on withdrawals due to adverse effects. In a general commentary, it is stated that "proportions of patients reporting adverse symptoms, including hypotensive symptoms, were similar in the 2 blood pressure groups". However, according to Table 5 in the same report, cough occurred more frequently in patients assigned to the lower blood pressure target (54.6% vs 47.0, p<0.05). 3.2.2.5 Number of antihypertensive drugs needed per patient Only the M D R D and the A A S K trials reported the number of drugs needed in each group in order to try to achieve the target blood pressure. As shown in Figure 12 and in Table 6, 0.5 more pharmacological antihypertensive agents were needed per patient in the groups randomized to the lower blood pressure target (p= 0.00001). Figure 12. Number of antihypertensive drugs needed per patient to achieve the target blood pressure Lower target Traditional target WMD Weight WMD Study n mean(sd) n mean(sd) (95%CI Random) % (95%CI Random) AASK Trial MDRD 540 432 3.04(1.14) 1.91(1.10) 554 408 2.39(1.18) 1.50(1.10) & 50.7 49.3 0.65(0.51,0.79) 0.41(0.26,0.56] Total(99%CI) 972 Test lor heterogeneity chi-square=5.39 df=1 p=0.02 Test for overall effect z=4.43 p=0.00001 962 • 100.0 0.53(0.22,0.84] •10 -5 0 5 10 Lower target Traditional target 55 The following information from the HOT trial (table 7) was estimated from a figure provided in an interim publication reporting results after 24 months of follow-up [64]. Due to these limitations, the data were not combined with the other trials. The estimated mean number of antihypertensive drugs per patient was 1.87 in the lower target groups combined (DBP < 80 mmHg and < 85 mmHg) and 1.68 in the traditional target group (DBP < 90 mmHg): p< 0.00001. The proportion of patients requiring high dose of one or more antihypertensive drugs (step 3 or higher), was larger in the lower target groups (44.9%) than in the traditional target group (31.4%): p< 0.0001. Table 7. Estimated percentage of patients by dose titration step according to target blood pressure group at 24 months of follow-up in the HOT trial Target Step 1 Step 2 Step 3 Step 4 Step 5 total (1 drug) (2 drugs) (2 drugs) (2 drugs) (3 drugs) < 80 mm Hg 24.6 25.9 20.5 10.4 18.2 99.6% < 85 mmHg 30.7 28.5 19.1 7.7 13.9 99.9% < 90 mmHg 40.0 28.4 16.1 6.2 9.1 99.8% 3.2.3 Summary When "lower" treatment target blood pressures (< 135/85 mmHg) are compared with "traditional" target blood pressures (< 140-160/90-100 mmHg), best available evidence from randomized controlled trials in the general population of patients with elevated blood pressure, including patients with diabetes or chronic renal disease, demonstrates that aiming for the lower target is associated with: a. No change in total, cardiovascular or non-cardiovascular mortality b. No data reported for the incidence of total serious adverse events 56 c. No change in the incidence of serious cardiovascular events, such as myocardial infarction, stroke, congestive heart failure, end-stage renal disease, or combined major cardiovascular events d. A lower achieved systolic blood pressure: 139.3 mmHg vs 143.2 mmHg, p= 0.003 e. A lower achieved diastolic blood pressure: 81.7 mmHg vs 85.1 mmHg, p< 0.00001 f. A larger number of antihypertensive drugs used per patient: 2.5 vs 2.0, p= 0.00001 g. A larger proportion of patients not achieving the blood pressure target h. No data reported for withdrawal rates due to adverse effects 57 4. D I S C U S S I O N 58 4.1 GENERAL DISCUSSION The objective in the treatment of patients with elevated blood pressure is to reduce morbidity and mortality and not simply to lower blood pressure. Many epidemiological studies have shown a continuous direct linear relationship between blood pressure and the incidence of cardiovascular events. The lower threshold for this relationship has not been established [5, 77]. More aggressive treatment in patients with elevated blood pressure aiming at lower blood pressure targets, assumes that lower blood pressure due to increased drug therapy achieves additional reductions in morbidity and mortality. However, these hypotheses cannot be used to make recommendations for increased drug therapy until they have been tested and validated by randomized controlled trials [78]. This systematic review and meta-analysis of randomized controlled trials summarizes the presently available appropriate trials testing blood pressure targets and includes almost 22,000 patients, with a mean follow-up period of 3.8 years. It includes all trials measuring clinical outcomes in patients randomized to "lower blood pressure targets" as compared with "traditional blood pressure targets" for the treatment of patients with elevated arterial blood pressure. The important observations from this systematic review are as follows. A l l of the identified trials assess diastolic or mean blood pressure targets. None of the trials compared different targets for systolic blood pressure. Therefore, at the present time we have no information regarding the benefits or harms of trying to achieve "lower targets" as compared with "traditional targets" for systolic blood pressure. Although mean arterial blood pressure (MAP), the parameter used as the comparison in some trials, is partially determined by systolic blood pressure, M A P cannot be used to estimate systolic blood pressure because every single value of M A P can be the result of many different combinations of systolic and diastolic blood pressure values. For example, using the proposed definition of M A P equal to 2/3 diastolic blood pressure plus 1/3 systolic blood pressure, a M A P equal to 107 mmHg could be the result of both 140/90 mmHg and 180/70 mmHg, among many other possible combinations of systolic/diastolic 59 values. The assumption that blood pressure measurements of 140/90 mmHg and 180/70 mmHg are the same based on the same estimated mean arterial pressure is likely to be wrong. This is especially so i f the observational data that pulse pressure is an independent and better predictor of cardiovascular risk than either systolic or diastolic blood pressure is confirmed [79-81]. Five of the 6 included RCTs compared diastolic or estimated diastolic BP targets of less than or equal to 85 mmHg with a diastolic or estimated diastolic BP target of less than or equal to 90 mmHg. In the remaining trial (Toto et al) the "traditional target" diastolic blood pressure was less than or equal to 95 mmHg, but still within the range of "traditional targets" specified in this systematic review. This trial was small and a sensitivity analysis excluding it does not change the results. Therefore, the results of this systematic review are generalizable to physicians prescribing drugs in an attempt to achieve diastolic targets of less than or equal to 85 mmHg as compared with a diastolic target of less than or equal to 90 mmHg. In patients randomized to the "lower targets" the mean systolic blood pressure was 4 mmHg lower and the mean diastolic blood pressure was 3 mmHg lower than patients randomized to the "traditional target" (139.3 mmHg vs 143.2 mmHg systolic, and 81.7 mmHg vs 85.1 mmHg diastolic respectively). However, despite these reductions in blood pressure, trying to achieve the "lower targets", as recommended by a recently published clinical guideline [31], instead of the "traditional target", did not result in any change in total, cardiovascular or non-cardiovascular mortality, and did not result in any change in cardiovascular or renal morbidity expressed as the incidence of myocardial infarction, stroke, congestive heart failure, the composite outcome of major cardiovascular events, or end-stage renal disease. Despite receiving a greater number of antihypertensive drugs, a larger proportion of patients did not achieve the lower target as compared with the traditional target. The reductions in blood pressure were not as great as the weighted mean difference in targets in the 6 trials, a reduction of 8.2 mmHg. This inability to achieve the difference in 60 target predominantly comes from the HOT trial and reflects a possible flaw in the design of that trial. By defining each group as less than or equal to a particular number, physicians tended to over-treat the higher target group, and were unable to achieve the targets in the lower target groups. If the targets were defined as a range of pressure for each group, a greater difference in blood pressure between the groups would most likely have been achieved. What is clear is that an increased number of antihypertensive drugs, and consequently an increase in inconvenience to the patient and economic costs to the patient or society, were required in order to achieve the lower blood pressure target. The lack of a difference in mortality and morbidity is robust as evidenced by the two most important measures: total mortality (RR=1.0, 95% CI 0.86 to 1.16) and total major cardiovascular events (RR- 0.95, 95% CI 0.84 to 1.08). Those numbers mean that we have acceptable confidence that the relative risk for total mortality is between a 14% decrease and a 16% increase, and that the relative risk for major cardiovascular events is between a 16% decrease and 8% increase. As can be seen, these ranges do not include the 20 to 25% difference that I would consider as being clinically significant. They do not, however, exclude the possibility of a small benefit or small harm associated with attempting to achieve lower diastolic blood pressure targets. The lack of a significant increase in total or cardiovascular mortality in the lower target group does not support the J curve hypothesis for diastolic BP targets < 85 mmHg in the general population of patients with elevated blood pressure. However, the distinct possibility exists that there is some harm associated with the lower targets, as the measures that would reflect this, total serious adverse events and withdrawals due to adverse effects, were not reported in these trials. One important argument against the conclusion that there is no difference in morbidity and mortality between the different targets is that the difference in the targets chosen in this review, 5 mmHg, was not great enough to show the predicted reduction in mortality and morbidity. This can be tested in this meta-analysis by including only those trials where the difference in the targets was at least 10 mmHg. Placing this restriction on the 61 review only excludes the middle group (< 85 mmHg) of the HOT trial. This re-analysis leads to a greater weighted mean difference in target diastolic blood pressure (10.3 mmHg), between the two groups. It also, as can be seen below, leads to a greater difference in the mean achieved blood pressures. However, there are no differences in mortality or morbidity (Table 8), and the conclusions remain the same as described above. Table 8.. Comparison of outcomes and achieved blood pressures in trials where the difference in the targets was at least 10 mmHg Outcome Number of Number of RR or W M D studies participants (99% CI) Total mortality 5 14647 1.00 (0.81, 1.25) C V mortality 4 14570 1.05 (0.76, 1.45) Non-CV mortality 4 14570 0.96 (0.71, 1.29) Myocardial infarction 3 13476 0.91 (0.68, 1.22) Stroke 3 13476 0.88 (0.62, 1.24) Congestive heart failure 2 950 1.06 (0.48, 2.32) Major C V events 4 14570 0.93 (0.77, 1.12) End-stage renal disease 2 1171 0.97 (0.68, 1.38) Achieved systolic BP 5 14647 -4.85 (-5.34, -4.36) Achieved diastolic BP 5 14647 -4.42 (-4.65, -4.20) The lack of randomized controlled trials designed to compare different targets for systolic blood pressure in this systematic review precludes making any conclusion regarding a systolic blood pressure target less than the "traditional targets" < 140 to 160 mmHg. Furthermore, the lack of a demonstrated therapeutic advantage and the missing information on safety do not support the strategy of trying to achieve any diastolic blood pressure target lower than the traditional target of less than or equal to < 90 mmHg in the general population of patients with elevated blood pressure. 62 In order to obtain generalizable results, meta-analysis combines all trials performed in the topic under study. However, looking globally at the general population of patients with elevated blood pressure included in this meta-analysis makes the assumption that there are no significant differences in blood pressure targets for all different subgroups of the populations included. As mentioned before, there are epidemiological and pathophysiological data suggesting that the concomitant presence of either diabetes mellitus or chronic renal disease increases the cardiovascular risk associated with elevated blood pressure. If patients with either diabetes mellitus or chronic renal disease have different risks and responses to lowering elevated blood pressure, they should be analyzed separately. Therefore, a sensitivity analysis performed in trials reporting outcomes in these subgroups of patients was performed. The results of the following sensitivity analyses are similar regardless of whether the range for lower targets chosen was <135/85 mmHg or < 130/80 mmHg. The range of < 130/80 mmHg was chosen for presentation for the following reasons: a. The differences in the achieved mean systolic (3.9 mmHg) and mean diastolic blood pressures (3.4 mmHg) between the lower and the traditional target groups were relatively small using <135/85 mmHg for the main analysis. Therefore, a wider separation between the targets might increase the chance of demonstrating a difference in outcomes between the groups. b. A target < 130/80 mmHg is the closest match with the targets defined in the individual trials analyzing outcomes in patients with diabetes or chronic renal disease. c. It has practical implications, because it allows evaluating the recommendations in current clinical guidelines for patients with diabetes and chronic renal disease. 63 4.2 SENSITIVITY ANALYSIS IN PATIENTS WITH DIABETES MELLITUS Diabetes was an exclusion criterion in the trial by Toto et al, and A A S K . In the M D R D trial, diabetic patients were excluded i f they required insulin; 25 of the included patients had Type 2 diabetes mellitus, but they could not be included in this analysis because their outcomes were not reported separately. Therefore, this sensitivity analysis is limited to two trials specifically carried-out in patients with diabetes mellitus, A B C D (H) and A B C D (N), and a sub-group analysis of the HOT trial, which separately reported some outcomes in patients with that disease (Table 9). Table 9. Outcomes in diabetic patients included in the HOT trial according to blood pressure target Outcome DBP < 80 mmHg DBP < 85 mmHg DBP < 90 mmHg (N= 499) (N= 501) (N= 501) Total mortality 17 29 30 C V mortality 7 21 21 Non-CV mortality 10 8 9 Myocardial infarction 7 8 14 Stroke 12 13 17 Major C V events 22 34 45 Mean (SD) achieved SBP 143.7(11.2) 146.4(11.4) 147.7(11.4) (mmHg) Mean (SD) achieved DBP 81.0 (5.0) 82.9 (4.7) 84.9 (5.6) (mmHg) Several issues regarding the design and the characteristics of these 3 trials must be mentioned before analyzing the results of the sensitivity analysis in diabetic patients. 64 Although not clearly specified, the subgroup analysis of diabetic patients in the HOT trial appears to be a post-hoc analysis, because it was not mentioned in any of the preliminary descriptions or reports of the trial [61-64]. In terms of baseline diastolic blood pressure, one of the trials, A B C D (N), included only normotensive diabetic patients, defined as a diastolic blood pressure between 80 and 89 mmHg. Twenty-six patients (5.4%) with isolated systolic hypertension (systolic blood pressure > 160 mmHg and diastolic blood pressure between 80 and 89 mmHg) were enrolled during the first year of recruitment, but none thereafter. On the other hand, both A B C D (H) and HOT only included patients with elevated blood pressure, but the criteria for inclusion was different. In A B C D (H), patients had a baseline diastolic blood pressure equal to or higher than 90 mmHg, whereas an inclusion criterion in HOT was a baseline diastolic blood pressure between 100 mmHg and 115 mmHg. It also must be noted that when A B C D (H) and A B C D (N) trials were combined, there was a difference in the proportion of patients with diagnosed cardiovascular or cerebrovascular disease at baseline randomized to either intensive or moderate blood pressure target (Table 10). This difference is clinically important and would be expected to have an impact on the outcomes, because having a greater proportion of patients with established vascular disease means that the higher target treatment group had a higher risk of future events than those assigned to the lower blood pressure target. Data from HOT was not included in Table 10 because it is not available. It can be assumed that baseline characteristics of diabetic patients in the HOT trial were similar between the target blood pressure groups, because randomization was blocked for several factors, including diabetes mellitus. However, there is no information in that respect. 65 Table 10. Distribution of patients with previous cardiovascular or cerebrovascular disease at baseline according to target blood pressure in the A B C D (H) and A B C D (N) trials Trial Lower target Traditional target P A B C D (H) 49% (116/237) 57% (133/233) 0.08 A B C D (N) 26% (62/237) 33% (80/243) 0.11 Total 37.5% (178/474) 44.7% (213/476) 0.02 Finally, it must be mentioned that these three trials were performed before the currently used diagnostic criteria for diabetes mellitus were established [82]. At that time, diabetes mellitus was defined as two fasting plasma glucose levels, measured on different days, higher than 7.7 mmol/L (140 mg/dL), instead of 7.0 mmol/L (126 mg/dL) as currently defined. The sensitivity analysis in patients with diabetes mellitus provided the following information, summarized in Table 11. Using the fixed effects model as the best estimate, the weighted mean achieved systolic blood pressure was 137.03 mmHg in the "lower target" group and 142.69 mmHg in the "traditional target" group giving the weighted mean difference shown in Table 11. The weighted mean achieved diastolic blood pressure was 78.52 mmHg in the lower target group and 84.02 mmHg in the traditional target group. The sensitivity analysis performed in diabetic patients demonstrates no statistically significant difference for any of the mortality and morbidity outcomes. However, in diabetics randomized to "lower target" (< 80 mmHg for diastolic blood pressure) as compared with a "traditional target" (< 90 mmHg) there is a trend towards a potentially clinically significant decrease in stroke, major cardiovascular events and total mortality. However, the results of this sensitivity analysis for cardiovascular events in diabetic patients do not achieve statistical significance for any outcome. In addition, for all of the 66 outcomes mentioned, the confidence intervals are very wide, demonstrating that very little confidence can be placed in the results of this sensitivity analysis and that more trials are required. Thus, despite achieving significantly lower mean systolic and mean diastolic blood pressures, the groups allocated to a target diastolic blood pressure < 80 mmHg did not achieve a statistically significant benefit in terms of cardiovascular morbidity or mortality compared with a DBP target < 90 mmHg. Moreover, as mentioned above the lack of reporting of total serious adverse events and withdrawals due to adverse effects prevents assessment of potential harm associated with the lower targets. Table 11. Comparison of outcomes and achieved blood pressures between "lower" and "traditional" blood pressure targets in patients with diabetes mellitus Outcome Number of Number of RR or W M D studies participants (99% CI) Total mortality 3 1950 0.63 (0.39, 1.01) C V mortality 3 1950 0.65 (0.20,2.13) Non-CV mortality 3 1950 0.62 (0.29, 1.29) Myocardial infarction 3 1950 0.98 (0.57, 1.69) Stroke 3 1950 0.64 (0.34, 1.23) Congestive heart failure 2 950 1.06 (0.48,2.32) Major C V events 3 1950 0.78 (0.56, 1.08) End-stage renal disease 0 0 Achieved systolic BP 3 1950 -5.66 (-6.99, -4.33) Achieved diastolic BP 3 1950 -5.50 (-6.09, -4.91) 67 4.3 SENSITIVITY ANALYSIS IN PATIENTS W I T H CHRONIC R E N A L DISEASE Although arbitrary, clinical trials have defined chronic renal disease based on a glomerular filtration rate lower than 70 mL/min/1.73m2 of body surface area. According to that parameter, three trials were included in the sensitivity analysis in patients with chronic renal disease: M D R D , Toto et al, and A A S K . Derangement in renal function was not an exclusion criterion in the HOT trial; however, it could not be included because there is no information regarding outcomes in patients with chronic renal disease included in that trial. A B C D (H) and A B C D (N) excluded patients receiving hemodialysis, peritoneal dialysis or had a serum creatinine concentration greater than 265 umol per liter (3 mg/dL). The information for many outcomes is scarce, because the reports focused mainly in surrogate markers of renal function. The great majority of information is derived from the A A S K trial. This sensitivity analysis provided the following information, summarized in Table 12. Table 12. Comparison of outcomes and achieved blood pressure between "lower" and "traditional" blood pressure targets in patients with chronic renal disease Outcome Number of Number of R R o r W M D studies participants (99% CI) Total mortality 2 1171 0.90 (0.52, 1.57) C V mortality 1 1094 0.90 (0.35, 2.27) Non-CV mortality 1 1094 0.87 (0.43, 1.78) Myocardial infarction 0 0 Stroke 0 0 Congestive heart failure 0 0 Major C V events 1 1094 0.87 (0.55, 1.39) End-stage renal disease 2 1171 0.97 (0.68, 1.38) Achieved systolic BP 2 1171 -12.69 (-14.52,-10.85) Achieved diastolic BP 2 1171 -6.91 (-8.02, -5.79) 68 Toto et al. and the A A S K trial reported achieved systolic and diastolic blood pressures. The M D R D trial was not included because it only reported achieved mean arterial pressure. Using the fixed effects model as the best estimate, the weighted mean systolic blood pressure was substantially lower, 128.2 mmHg, in the "lower target" group than in the "traditional target" group, 140.9 mmHg. Likewise, the achieved diastolic blood pressure was significantly lower in the lower target group: 78.3 mmHg vs 85.2 mmHg . As mentioned above (Figure 12), 0.5 more pharmacological antihypertensive agents were needed per patient in the groups with chronic renal disease randomized to the lower blood pressure target. Although it is a surrogate marker, it is worthwhile to look at mean decline in glomerular filtration rate (GFR), because it was the main objective of the three trials included in this sensitivity analysis, and also because preserving renal function is promoted as one of the main benefits from lower target blood pressures in patients with chronic renal disease. In M D R D , the standard deviation of decline in GFR was reported for the entire trial, not annually; for that reason, it was imputed using the weighted mean SD from the other two trials. As shown in Figure 13, there was no difference in mean annual decline in GFR (expressed in ml/min/1.73m2 per year) between patients randomized to lower as compared with traditional blood pressure targets (p= 0.7). Figure 13. Mean decline in glomerular filtration rate (ml/min/ 1.73m2 per year) according to blood pressure target in patients with chronic renal disease Lower target Traditional target WMD Weight WMD Study n •nean(sd) n mean(sd) (95%CI Fixed) % (95%CI Fixed) AASK Trial 540 2.21(3.95) 554 1.95(4.00) I 52.0 0.26|-0.21,0.73] MDRD 432 3.56(3.88) 408 4.10(3.93) m 41.3 -0.54[-1.07,-0.01] Toto 42 0.31(2.92) 35 0.05(2.96) 6.6 0.26[-1.06,1.58] Total(99%CI) 1014 997 < • 100.0 -0.07[-0.52,0.38] Test for heterogeneity chi-square=S .16 df=2 p=0.076 Test for overall effect z=0.41 p=0.7 -10 -5 0 5 10 Lower target Traditional target 69 The sensitivity analysis in patients with elevated arterial blood pressure and chronic renal disease did not show any statistically reduced mortality, total cardiovascular events or end stage renal disease with a lower diastolic blood pressure target (< 80 mmHg) mmHg as compared with a traditional diastolic blood pressure target (< 90 mmHg). The absence of data precludes any conclusion regarding other specific cardiovascular serious adverse events. Even the surrogate outcome decline in GFR was not reduced by trying to achieve lower blood pressure targets. These findings were demonstrated despite achieving substantially lower blood pressures in these trials in the lower target group. The safety of the more intensive treatment cannot be assessed due to the lack of information regarding total serious adverse events and withdrawals due to adverse effects. 4.4 LIMITATIONS The main limitations in this meta-analysis are as follows. The possibility of investigator bias is a limitation, because the strategy of trying to achieve a target blood pressure does not allow performing a double-blind trial. Open trials have lower ability to evaluate both, efficacy and safety outcomes, as compared with double-blind trials. This was dealt with in 4 of the trials by ensuring that the investigators performing the outcome evaluation were blinded to the randomized group [83]. This approach was described in the design of the 4 more recent trials, but not in the two earlier trials (MDRD and Toto). To further decrease investigator bias, it is also very important to strictly define the outcomes. In that regard, it is noteworthy that in the HOT trial the Clinical Event Committee rejected 24% of all investigator-reported events. On the other hand, a randomized open trial with blinded end-point evaluation has the advantage of greater resemblance to common clinical practice. Another limitation of this meta-analysis is the fact that one single study provides the great majority of participants included. There has been some concern regarding the convenience of performing a meta-analysis in those cases, especially when the results in 70 the meta-analysis and the mega-trial have been contradictory [84]. This is not the case in this meta-analysis since the results are similar whether or not the largest trial is included. The differences between trials in the definition of outcomes are a limitation. For example, the HOT trial did not include congestive heart failure in the definition of major cardiovascular events. HOT, but not the remaining trials, reported also silent myocardial infarctions. The absence of a statistically significant difference in the incidence of myocardial infarctions and in major cardiovascular between "lower" and "traditional" targets found in this meta-analysis would have been smaller i f silent infarcts were included. Another limitation is the definition of targets based on mean arterial pressure in some trials. This strategy allowed the trialist to make decisions based on a single value which reflected both diastolic and systolic blood pressure. However, it made it difficult to be certain that the comparisons with other trials using a diastolic blood pressure target were the same. Some trials are lacking data. There is very little information on several outcomes, especially those not reported in the HOT trial. The information on mortality and on end-stage renal disease from the M D R D trial could not be included for the reasons already mentioned. In spite of specific requests to the trials' main authors, the missing information was not provided. Hence, reporting bias is another limitation. As in most of the trials analyzing clinical outcomes in patients treated for elevated blood pressure, a major limitation in this analysis is to be certain whether the achieved blood pressures accurately reflect the real differences in blood pressure over the duration of the trials. This approach does not take into account the great variability occurring in that parameter within and between days [85]. The trial reports were not very clear as to the times that the blood pressures were measured and how that was standardized across patients in the trial. 71 Finally, the comparison between different targets assumes that the different drug treatments used to achieve those targets did not lead to different treatment effects between the lower and traditional targets. 4.5 COMPARISON WITH OTHER PUBLISHED REPORTS 4.5.1 Clinical guidelines The results and conclusions of this meta-analysis differ from recommendations provided by several currently used clinical guidelines for the treatment of high blood pressure. It must be mentioned that those recommendations are based either on the results of observational cohort studies or on the interpretation of results in randomized controlled trials (RCT) as i f they were observational studies in relation to achieved blood pressure, instead of according to the randomized groups. Due to the common misunderstanding of this issue, it deserves a more detailed explanation. Although observational studies can provide very interesting and useful information, it is well established that data obtained from these studies has several limitations [78]. Observational studies are subject to confounding due to the participation of other variables, and also to selection bias, which means that there is differential assignment of subjects to treatment groups based on subject characteristics [86, 87]. Therefore, observational studies must be limited to generation of hypotheses that can validated in properly designed, conducted and reported RCTs, which are the gold standard for establishing evidence of causal associations in medical investigation [86, 87]. The interpretation of results in randomized controlled trials (RCT) in relation to achieved blood pressure has been used in the HOT trial. This trial was designed as a randomized trial to measure outcomes based on 3 different target diastolic blood pressures. Mortality 72 and cardiovascular morbidity were not different in the 3 different targets groups. However, instead of making conclusions based on the randomized trial as designed, the authors did a further analysis of the trial as i f it was a prospective observational study and made their conclusions based on that analysis. In this design they combined all randomized groups into one and reported outcomes based on the blood pressures achieved during follow-up. That approach is not appropriate, because the main objective of randomization, to assure a balanced distribution of all baseline patient conditions and characteristics, is lost [88]. The following explanations may help to better understand this important concept. There is no clear cause identified for more than 90% of cases of elevated blood pressure. Many different genetic, neurohormonal and environmental factors, as well as functional and structural alterations in vascular tissues, have been implicated in the pathogenesis of elevated blood pressure [89]. Thus, rather than being a specific and unique condition, it is more likely a single manifestation common to many different conditions. If analysis is performed according to achieved blood pressure, patients in whom the pathogenetic mechanism is easiest to control and hence have the greatest decrease in blood pressure in response to antihypertensive therapy are being selected. These individuals may have less likelihood of developing cardiovascular events independent of therapy. Achieved blood pressure is likely related to baseline pressure. That is, the lower the baseline diastolic blood pressure, the less absolute reduction needed to achieve a low diastolic blood pressure. It is probable that patients with lower baseline blood pressure have a greater probability of being included in the lower achieved blood pressure group and vice versa for patients with a higher baseline blood pressure. A lower achieved blood pressure is also likely the result of a greater compliance with both, pharmacological and non-pharmacological treatment. Non-pharmacological therapy, which includes life-style changes such as a healthy diet, physical exercise, weight reduction, cessation of smoking, etc. can provide additional cardiovascular and non-cardiovascular benefits independent of the effect on blood pressure. 73 Lack of tolerability or responsiveness to antihypertensive treatment could be related to an increased incidence of cardiovascular events. Therefore, those patients with higher achieved blood pressures would be expected to have worse outcomes independent of treatment. In other words, selection bias is automatically introduced when outcomes are analyzed in groups different than originally established by random assignment. As can be seen, analyzing the results of a RCT by achieved blood pressure has the same limitations, selection bias and confounding, as an observational study. In fact the results reported in the HOT trial on achieved blood pressure are exactly what would be predicted from observational data, patients with lower blood pressures have less adverse cardiovascular outcomes than patients with higher pressures. Therefore, analysis of outcomes according to achieved blood pressure is not appropriate, provides misleading information, and thus must not be encouraged. The limitations previously mentioned are prevented when results from a RCT are analyzed according to intention-to-treat principle, such as used in this meta-analysis, because then the only difference between groups is the treatment strategy under consideration, and consequently any difference in outcomes can be attributed to the allocated treatment strategy [90]. 4.5.2 Blood Pressure Lowering Treatment Trialists' Collaboration meta-analysis The results of this systematic review differ with those reported in another recent meta-analysis performed by the Blood Pressure Lowering Treatment Trialists' Collaboration [76]. In their study, the BPLTTC compared "less intensive" with "more intensive" blood pressure lowering regimens, without defining any specific blood pressure target values 74 for that grouping. The BPLTTC analysis included the UKPDS-38 trial, which was excluded in the present review. The BPLTTC meta-analysis showed a decreased incidence in stroke and major cardiovascular events with "more intensive" blood-pressure-lowering regimens compared with "less intensive" regimens. It showed no significant difference in the incidence of coronary heart disease, congestive heart failure, death due to cardiovascular causes, or total mortality. However, from detailed information about the trials and analysis performed, available at the BPLTTC website (http://www.iih.org/bplttc') it is clear that for both outcomes, stroke and major cardiovascular events, the differences obtained were due to the inclusion of results from the UKPDS-38 trial. UKPDS 38 only included patients with type 2 diabetes mellitus, and extrapolating findings from diabetic patients to the whole population may be misleading. As mentioned before, the reasons for excluding UKPDS 38 in this review were two-fold. Firstly, the target for the "low target" group was < 150/85 mmHg. This is correct for the diastolic but is in the mid-range for the "traditional target" for systolic blood pressure. Secondly, it included an inappropriately high blood pressure target in the less intensive treatment group: initially the target was < 200/105 mmHg, and after 5 years it was reduced to < 180/105 mmHg. Furthermore, given that the mean baseline blood pressure in that group was 160/94 mmHg, setting such a high target implies that many patients in that group did not receive any antihypertensive therapy in spite of having blood pressure values above which treatment has been proven to be beneficial. In fact, the trial report states that in the group assigned to less tight blood pressure control, "patients did not take any antihypertensive treatments for 43% of the total person years" (UKPDS 38). So, when analyzing the results, it must be clear that rather than comparing two blood pressure treatment targets, UKPDS 38 is primarily a comparison between treatment to a traditional target, < 150/85 mmHg, and no treatment. For these reasons, it is clear that this trial does not provide any information relevant to the comparison between "lower" and "traditional" blood pressure treatment targets. 75 5. C O N C L U S I O N 76 5.1 CONCLUSIONS The conclusions of this systematic review have implications for both clinical practice and further research. 5.1.1 Implications for clinical practice 1. The optimal target for systolic blood pressure has not been tested in randomized controlled trials. This is particularly important because current trends in clinical practice are to focus more on the control of elevated systolic blood pressure. In the absence of evidence, systolic blood pressure targets must be those that have been demonstrated to be better than placebo or no treatment in randomized controlled trials, that is, < 150 to 160 mmHg. The lower of this systolic target range also applies to diabetics as that was the systolic target used in the UKPDS trial in diabetics. It should be noted that in these trials 30 to 40% of patients were not able to achieve the specified target despite maximal utilization of the antihypertensive drugs allowed. 2. This systematic review primarily analyzed diastolic blood pressure targets. The 6 RCTs identified adequately tested whether "lower targets" reduced morbidity and mortality as compared with "traditional targets". Despite achieving lower blood pressures and requiring a greater number of antihypertensive drugs, treating patients with elevated blood pressure to targets < 85 mmHg, was not associated with a reduction in mortality or morbidity when compared with the traditional target, < 90 mmHg. The same conclusion is true i f one compares targets < 80 mmHg with < 90 mmHg. Therefore, for the general population of patients with elevated blood pressure, the diastolic target should be < 90 mmHg and there is no justification for attempting to treat to lower targets. 77 3. For the population of diabetic patients, the diastolic blood pressure target should be < 90 mmHg. However, the numbers of events and patients in this subgroup is small and therefore the confidence intervals are much wider. Because of this, the possibility exists that a clinically significant benefit for lower targets could have been missed. That is particularly so for the outcomes total mortality, major cardiovascular events, and stroke. This observation makes the need for an RCT designed to answer that question in diabetics rather urgent (see below under research implications). 4. For the population of patients with chronic renal disease, the diastolic blood pressure target should be < 90 mmHg. However, because of the lack of data and the small number of events and patients in this subgroup, the possibility exists that a clinically significant benefit or harm could have been missed. The lack of any benefit for "lower targets" in terms of the renal outcomes, end-stage renal disease and decline in glomerular filtration rate calls into question the guidelines recommending diastolic blood pressure targets lower than traditional targets in this subgroup. 5. A l l trials in this meta-analysis failed to report the outcomes that would reflect harm associated with "lower targets", non-fatal serious adverse events and withdrawals due to adverse effects. Therefore there is a distinct possibility that harm associated with "lower targets" will be demonstrated, when this data is made available. 6. Conclusions and recommendations for blood pressure targets based on blood pressures achieved in randomized controlled trials are misleading and should be discouraged. 5.1.2 Implications for research 1. Because guidelines are presently recommending unproven low systolic blood pressure targets of < 130 mmHg in patients with diabetes mellitus, a trial testing 78 systolic blood pressure targets in this population is urgently needed. Important components of such a trial are listed below. Methods and design: Open label randomized controlled trial Standardized automated blood-pressure-measuring equipment to measure the average of 3 blood pressures in a rested seated position. Standardized time of measurement in relation to drug intake (2-12 hours post drug). Follow-up for at least 5 years. - An independent clinical outcome committee, blinded to the assigned target strategy. Patients: Adult ambulatory patients with type 2 diabetes mellitus and systolic blood pressure > 150 mmHg on no antihypertensive treatment for 4 to 6 weeks. - A sample size of 3500 subjects will be required to detect a 20% reduction in the primary outcome specified with a significance level of a= 0.05, and a power of P= 0.80. Interventions: - Patients randomly allocated to two different systolic blood pressure target ranges 120 to 134 mmHg and 135 to 149 mmHg. . - Medication adjustments only allowed i f average blood pressure measurements on 2 occasions at least one week apart are outside the range. If these confirmed measurements are above the range, medication is added and i f below the range medication is reduced. - A standardized stepped care-approach to drug treatment. 1. Hydrochlorothiazide 12.5 mg once daily to start. 2. Increase the dose of hydroclorothiazide to 25 mg once daily. 3. Add a starting dose of a once daily beta-adrenergic blocker 79 4. Double the dose of the beta-adrenergic blocker drug 5. Add a starting dose of a once daily A C E inhibitor. 6. Double the dose of the A C E inhibitor 7. Add a starting dose of a once daily calcium channel blocker 8. Double the dose of the calcium channel blocker. d. Outcomes: The primary outcome is total cardiovascular events defined as fatal and non-fatal myocardial infarction, plus fatal and non-fatal strokes, plus hospitalization due to heart failure, plus all other cardiovascular deaths. Secondary outcomes will be total mortality, total non-fatal serious adverse events and withdrawals due to adverse events. - The analysis will be performed by the intention-to-tfeat principle. A similar trial design would also be appropriate for the subgroup patients with elevated blood pressure and chronic renal disease and for other subgroups of patients with elevated blood pressure. 80 BIBLIOGRAPHY Bibliography 1- Stokes J, Kannel WB, Wolf PA, et al. The relative importance of selective risk factors for various manifestations of cardiovascular disease among men and women from 35 to 64 years. 30 years of follow-up in the Framingham Study. Circulation 1987;75 (Suppl V):V65-V73. 2- Stamler J, Stamler R, Neaton JD. BP, systolic and diastolic and cardiovascular risk. Arch Intern Med 1993;153:598-615. 3- Kannel WB. Blood pressure as a cardiovascular risk factor: prevention and treatment. J A M A 1996;275:1571-6. 4- Chalmers J et al. WHO-ISH Hypertension Guidelines Committee. 1999 World Health Organization-International Society of Hypertension Guidelines for the management of hypertension. J Hypertens 1999;17:151-85. 5- MacMahon SW, Peto R, Cuttler J, et al. Blood pressure, stroke and coronary heart disease. Part 1. Prolonged differences in blood pressure: prospective observational studies collected for the regression dilution bias. Lancet 1990;335:765-74. 6- Veterans Administration Cooperative Study Group on Antihypertensive Agents. Effects of treatment on morbidity in hypertension. Results in patients with diastolic blood pressures averaging 115 through 129 mmHg. J A M A 1967, 202:116-22. 7- Veterans Administration Cooperative Study Group on Antihypertensive Agents. Effects of treatment on morbidity in hypertension. II. Results in patients with diastolic blood pressure averaging 90 through 114 mmHg. J A M A 1970;213:1143-52. 8- Wright J M , Lee C H , Chambers GK. Systematic review of antihypertensive therapies: does the evidence assist in choosing a first-line drug? C M A J 1999;161:25-32. 9- Wolff FW, Lindeman RD. Effects of treatment in hypertension. Results of a controlled study. JChronDis 1966;19:227-40. 10- Perry H M . Treatment of mild hypertension. Preliminary results of a two-year feasibility trial. Veterans Administration-National Heart, Lung, and Blood Institute Study Group for Cooperative Studies on Antihypertensive Therapy: mild hypertension. CircRes 1977;40(1 Suppl): 180-7. 11- Carter A B . Hypotensive therapy in stroke survivors. Lancet 1970;1:485-9. 82 12- Barraclough M , Joy M D , MacGregor GA, et al. Control of moderately raised blood pressure. Report of a co-operative randomized controlled trial. B M J 1973;3:434-6. 13- Australian Therapeutic Trial in Mild Hypertension. Report by the management committee. Lancet 1980;1:1261-7. 14- Kuramoto K , Matsushita S, Kuwajima I, Murakami M . Prospective study on the treatment of mild hypertension in the aged. Jpn Heart J 1981 ;22:75-85. 15- Amery A , Brixko P, Clement D, et al. Mortality and morbidity results from the European Working Party on High Blood Pressure in the Elderly Trial. Lancet 1985;1:1349-54. 16- Helgeland A. Treatment of mild hypertension: a five year controlled drug trial. The Oslo Study. A m J Med 1980;69:725-32. 17- Medical Research Council Working Party. M R C trial of treatment of mild hypertension: principal results. B M J 1985;291:97-104. 18- M R C Working Party. Medical Research Council trial of treatment of hypertension in older adults: principal results. B M J 1992;304:405-12. 19- SHEP Cooperative Research Group. Prevention of stroke by antihypertensive drug treatment in older persons with isolated systolic hypertension. Final results of the Systolic Hypertension in the Elderly Program (SHEP). J A M A 1991;265:3255-64. 20- Staessen JA, Fagard R, Thijs L, et al. Randomised double-blind comparison of placebo and active treatment for older patients with isolated systolic hypertension. Lancet 1997;350:757-64. 21- Wright J M , Musini V . Personal communication. 22- Laurent S. Guidelines from the British Hypertension Society. The lower the pressure the better. B M J 2004;328:593-4. 23- American Diabetes Association. Standards of medical care in diabetes. Diabetes Care2004;27 (Suppl 1):S15-S35. 24- LusisAJ. Atherosclerosis. Nature 2000;407:233-41. 25- Glass CK, Witztum JL. Atherosclerosis: the road ahead. Cell 2001;104:503-16. 26- Plutzky J, Viberti G, Haffner S. Atheroslerosis in type 2 diabetes mellitus and insulin resistance: mechanistic links and therapeutic targets. J Diab Compl 2002;16:401-15. 83 27- Garg J, Bakris GL. Treatment of hypertension in patients with renal disease. Cardiovasc Drug Ther 2001;16:503-10. 28- Chobanian A V , Bakris GL, Black HR, et al. The seventh report of the Joint National Committee on prevention, detection, evaluation, and treatment of high blood pressure. The JNC 7 report. J A M A 2003;289:2560-72. 29- Guidelines Committee. 2003 European Society of Hypertension - European Society of Cardiology guidelines for the management of arterial hypertension. J Hypertens 2003;21:1011-53. 30- World Health Organization, International Society of Hypertension Writing Group. 2003 World Health Organization (WHO/Intemational Society of Hypertension (ISH) statement on management of hypertension. J Hypertens 2003;21:1983-92. 31- Williams B, Poulter NR, Brown MJ , et al. British Hypertension Society guidelines for hypertension management 2004 (BHS-IV): summary. B M J 2004;328:634-40. 32- Farnett L, Mulrow CD, Linn WD, et al. The J-curve phenomenon and the treatment of hypertension. J A M A 1991;265:489-95. 33- Cruickshank JM. Coronary flow reserve and the J-curve relation between diastolic blood pressure and myocardial infarction. Br Med J 1988;297:1227-30. 34- Voko Z, Bots M L , Hofrnan A , et al. J-shaped relation between blood pressure and stroke in treated hypertensives. Hypertension 1999;34:1181-7. 35- Zanchetti A , Hansson L, Clement D, et al on behalf of the HOT Study Group. Benefits and risks of more intensive blood pressure lowering in hypertensive patients of the HOT study with different risk-profiles: does a J-shaped curve exist in smokers? J Hypertens 2003;21:797-804. 36- Lund-Johansen P. Intensive blood pressure treatment: beneficial for all but the smoking hypertensives? J Hypertension 2003;21:697-700. 37- Mulrow CD. Systematic reviews: rationale for systematic reviews. B M J 1994;309:597-9. 38- Bailar JC. The practice of meta-analysis. J Clin Epidemiol 1995;48:149-57. 39- Cook DJ, Mulrow C, Haynes RB. Systematic reviews: synthesis of best evidence for clinical decisions. Ann Intern Med 1997;126:376-80. 84 40- Lau J, Ioannidis JPA, Schmid CH. Summing up evidence: one answer is not always enough. Lancet 1998;351:123-7. 41- Alderson P, Green S, Higgins JPT, editors. Cochrane Reviewers' Handbook 4.2.2. [updated March 2004]. In: The Cochrane Library, Isue 1, 2004. Chichester, U K : John Wiley & Sons, Ltd. 42- Lau J, Ioannidis JPA, Schmid CH. Quantitative synthesis in systematic reviews. Ann Intern Med 1997;127:820-6. 43- Whitehead A , Whitehead J. A general parametric approach to the meta-analysis of randomized clinical trials. Stat Med 1991; 10:1665-77. 44- Arauz-Pacheco C, Parrott M A , Raskin P. The treatment of hypertension in adult patients with diabetes. Diabetes Care 2002;25:134-47. 45- Aung K, Htay T. Isolated systolic hypertension in the elderly. Cardiovasc Rev Rep 2003;24:423-30. 46- Cruickshank J M . Antihypertensive treatment and the J-curve. Cardiovasc Drug Ther 2000;14:373-9. 47- Hansson L. Antihypertensive treatment: does the J-curve exist? Cardiovasc Drug Ther 2000;14:367-72. 48- Hansson L. How far should we lower blood pressure in the elderly. Cardiovasc Drug Ther 2001;15:275-9. 49- Izzo JL, Levy D, Black HR. Importance of systolic blood pressure in older Americans. Hypertension 2000;35:1021-4. 50- Prisant L M . Diabetes mellitus and hypertension: a mandate for intense treatment according to new guidelines. A m J Therap 2003;10:363-9. 51- Snow V , Weiss K B , Mottur-Pilson C. The evidence base for tight blood pressure control in the management of type 2 diabetes mellitus. Ann Intern Med 2003;138:587-92. 52- Vijan S, Hayward R A . Treatment of hypertension in type 2 diabetes mellitus: blood pressure goals, choice of agents, and setting priorities in diabetes care. Ann Intern Med 2003;138:593-602. 85 53- Zanchetti A , Ruilope L M . Antihypertensive treatment in patients with type-2 diabetes mellitus: what guidance from recent controlled randomized trials? J Hypertens 2002;2099-l 10. 54- JNV-6 1997. The sixth report of the Joint National Committee on prevention, detection, evaluation, and treatment of high blood pressure. Arch Intern Med 1997;157:2413-45. 55- McAlister F A , Zarnke K B , Campbell N R C et al. The 2001 Canadian recommendations for the management of hypertension: Part two - Therapy. Can J Cardiol 2002;18:625-41. 56- Klahr S, Levey AS, Beck GJ, et al for the Modification of Diet in Renal Disease Study Group. N Eng J Med 1994;330:877-84. 57- Peterson JC, Adler S, Burkart J M , et al for the Modification of Diet in Renal Disease (MDRD) Study Group. Ann Intern Med 1995;123:754-62. 58- Lazarus J M , Bourgoignie JJ, Buckalew V M , et al for the Modification of Diet in Renal Disease Study Group. Hypertension 1997;29:641-650. 59- Toto RD, Mitchell HC, Smith RD, et al. "Strict" blood pressure control and progression of renal disease in hypertensive nephrosclerosis. Kidney Int 1995;48:851-9. 60- Hansson L , Zanchetti A , Carruthers SG, et al for the HOT Study Group. Effects of intensive blood-pressure lowering and low-dose aspirin in patients with hypertension: principal results of the Hypertension Optimal Treatment (HOT) randomized trial. Lancet 1998;351:1755-62. 61- Hansson L, for the HOT Study Group. The Hypertension Optimal Treatment Study (The HOT Study). Blood Pressure 1993;2:62-8. 62- Hansson L, Zanchetti A , for the HOT Study Group. The Hypertension Optimal Treatment (HOT) Study. Patients characteristics, randomization, risk profiles, and early blood pressure results. Blood Pressure 1994;3:322-7. 63- Hansson L, Zanchetti A , for the HOT Study Group. The Hypertension Optimal Treatment (HOT) Study: 12-month data on blood pressure and tolerability with special reference to age and gender. Blood Pressure 1995;4:313-9. 86 64- Hansson L , Zanchetti A , for the HOT Study Group. The Hypertension Optimal Treatment (HOT) Study: 24-month data on blood pressure and tolerability. Blood Pressure 1997;6:313-7. 65- Hansson L. The Hypertension Optimal Treatment study and the importance of lowering blood pressure. JHypertens 1999;17(suppl 1):S9-S13. 66- Zanchetti A , Hansson L, Dahlof B, on behalf of the HOT Study Group. Effects of individual factors on the incidence of cardiovascular events in the treated hypertensive patients of the Hypertension Optimal Treatment Study. J Hypertens 2001;19:1149-59. 67- Estacio RO, Jeffers BW, Hiatt WR, et aj. The effect of nisoldipine as compared with enalapril on cardiovascular outcomes in patients with non-insulin-dependent diabetes and hypertension. N Eng J Med 1998;338:645-52. 68- Estacio RO, Gifford N , Jeffers BW, Schrier RW. Effect of blood pressure control on diabetic microvascular complications in patients with hypertension and type 2 diabetes. Diabetes Care 2000;23 (suppl 2): B54-B64. 69- Schrier RW, Estacio RO, Esler A , Mehler P. Effects of aggressive blood pressure control in normotensive type 2 diabetic patients on albuminuria, retinopathy and strokes. Kidney Int 2002;61:1086-97. 70- Wright JT, Bakris G, Greene T, et al for the African American Study of Kidney Disease and Hypertension Study Group. Effect of blood pressure lowering and antihypertensive drug class on progression of hypertensive kidney disease. Results from the A A S K Trial. J A M A 2002;288:2421-31. 71- The B B B Study Group. The B B B study: a prospective randomized study of intensified antihypertensive treatment. J Hypertens 1988;6:693-7. 72- Hansson L for the B B B Study Group. The B B B Study Group: the effect of intensified antihypertensive treatment on the level of blood pressure, side-effects, morbidity and mortality in "well-treated" hypertensive patients. Blood Pressure 1994;3:248-54. 73- Hypertension in Diabetes Study Group. Hypertension in diabetes study IV. Therapeutic requirements to maintain tight blood pressure control. Diabetologia 1996;39:1554-61. 87 74- U K Prospective Diabetes Study Group. Tight blood pressure control and risk of macrovascular complications n type 2 diabetes. B M J 1998;317:703-13. 75- Lewis JB, Berl T, Bain RP, et al. Effect of intensive blood pressure control on the course of type 1 diabetic nephropathy. A m J Kidney Dis 1999;34:809-17. 76- BPLTTC. Blood Pressure Lowering Treatment Trialists' Collaboration. Effects of different blood-pressure-lowering regimens on major cardiovascular events: results of prospectively-designed overviews of randomized trials. Lancet 2003;362:1527-35. 77- 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;360:1903-13. 78- Gueyfiier F. Observational epidemiological studies: values and limitations. J Hypertens 2001;21:673-5. 79- Franklin SS, Khan SA, Wong ND, et al. Is pulse pressure more important than systolic blood pressure in predicting coronary heart disease events? Circulation 1999;100:354-60. 80- Millar JA, Lever A F , Burke A . Pulse pressure as a risk factor for cardiovascular events in the M R C Mild Hypertension Trial. J Hypertens 1999;17:1065-72. 81- Benetos A , Zurek M , Morcel J, et al. A decrease in diastolic blood pressure combined with an increase in systolic blood pressure is associated with a higher cardiovascular mortality in men. J A m Coll Cardiol 2000;35:673-80. 82- American Diabetes Mellitus 1997. The Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care 1997;20:1183-97 83- Hansson L, Hedner T, Dahlof B. Prospective randomized open blinded end-point (PROBE) study. A novel design for intervention trials. Blood Pressure 1992;1:113-9. 84- Borzak S, Ridker P M . Discordance between meta-analyses and large-scale randomized, controlled trials. Examples from the management of acute myocardial infarction. Ann Intern Med 1995;123:873-7. 85- Parati G, Ulian L, Santucciu C, et al. Blood pressure variability, cardiovascular risk and antihypertensive treatment. J Hypertens 1995;13(suppl 4):S27-S34. 88 86- Katz DL. Clinical epidemiology and evidence-based medicine: Fundamental principles of clinical reasoning and research. Sage Publications; 2001. 87- MacMahon S, Collins R. Reliable assessment of the effects of treatment on mortality and major morbidity, II: observational studies. Lancet 2001;357:455-62. 88- Kunz R, Oxman A D . The unpredictability paradox: review of empirical comparisons of randomized and non-randomised clinical trials. B M J 1998;317:1185-90. 89- Oparil S, Zaman A , Calhoun DA. Pathogenesis of hypertension. Ann Intern Med 2003;139:761-76. 90- Peto R, Collins R, Gray R. Large-scale randomized evidence: large simple trials and overviews of trials. J Clin Epidemiol 1995;48:23-40. 89 7. APPENDIX 90 V I ,t> Ovid M E D L I N E ( R ) <1966 to April Week 4 2004> Help # Search History Results Display 1 randomized controlled trial.pt. 187799 Display 2 randomized controlled trials.mp. 34057 Display 3 randomized controlled trial.mp. 7147 Display 4 controlled clinical trial.pt. 66099 Display j 5 controlled clinical trials.mp. 6035 Display 6 controlled clinical trial.mp. 3670 Display 7 random allocation.mp. 50949 Display 8 or/1-7 314531 Display 9 exp animal/ 3635804 Display 10 8 not 9 290901 Display 11 clinical trial.pt. 380263 Display 12 clinical trials.mp. 150842 Display 13 clinical trial.mp. 29838 Display ^ 14 exp clinical trials/ 153255 Display 15 (clin$ adj25 trial$).mp. 98621 Display 16 random$.mp. 324031 Display ; 17 exp research design/ 178686 Display 18 research design.mp. 42813 Display 19 or/11-18 693287 Display 20 19 not 9 608880 Display 21 20 not 10 317979 - • Display 22 comparative studies.mp. 10613 Display 23 comparative study.mp. 1121039 'Display-'-91 25 evaluation studies.mp. 115506 Display ! 26 evaluation study.mp. 1129 Display 27 follow up studies.mp. 285039 Display 28 follow up study.mp. 16671 Display 29 prospective studies.mp. 177463 Display 30 prospective study.mp. 47468 Display 31 (controls or prospective$ or volunteer$).mp. 1482763 Display 32 or/22-31 2881270 Display j 33 32 not 9 2013834 Display 34 33 not (10 or 21) 1598503 Display 35 10 or 21 or 34 2207383 Display 36 arterial hypertension.mp. 13795 Display 37 hypertension.mp. 198170 Display 38 high blood pressure.mp. 5359 Display j 39 elevated blood pressure.mp. 2396 Display ; 40 hypertensive patients.mp. 14365 41 or/36-40 200984 Display | 42 target level.mp. 298 Displ ay 43 target blood pressure.mp. 206 Display 44 target systolic blood pressure.mp. 10 Display 45 target diastolic blood pressure.mp. 29 Display 46 intensive treatment.mp. 1884 Display 47 intensive blood pressure treatment.mp. 1 Display 48 intensive antihypertensive treatment.mp. 16 Display 49 intensive control.mp. 94 ] Di splay v 50 intensive blood pressure control.mp. 13 Display 51 tight control.mp. 567 ' Display 52 tight blood pressure control.mp. 43 Display 53 strict control.mp. 612 Display | 92 55 or/42-54 3760 Display j 56 41 and 55 505 Display O V I D E M B A S E <1980 to 2004 Week 19> # Search History Results Display 1 randomized controlled trial.pt. 0 -2. randomized controlled trials.mp. 3904 Display j 3 randomized controlled trial.mp. 87082 Display •4 controlled clinical trial.pt. 0 -5 - controlled clinical trials.mp. 3501 pDJ7p!an 6 controlled clinical trial.mp. 3316 f f > ; 6 i s p l a y ? ! j 7 random allocation.mp. 467 D.splay 8 or/1-7 94780 Display 9 exp animal/ 77770 Display 10 8 not 9 94719 Display > 11 clinical trial.pt. 0 -12 clinical trials.mp. 48026 Display | 13 clinical trial.mp. 308797 Display 14 exp clinical trials/ 304379 Display 15 (clin$ adj25 trial$).mp. 91703 Display ] 16 random$ .mp. 273040 17 exp research design/ 843746 display ~\ 18 research design.mp. 4402 Display 19 or/11-18 1191217 Display , 20 19 not 9 1184755 Display 21 20 not 10 1090036 Display 22 comparative studies.mp. 6452 Display 23 comparative study.mp. 70671 Display 24 exp evaluation studies/ 24221 Display | 25 evaluation studies.mp. 678 [ Display 26 evaluation study.mp. 968 Display 27 follow up studies.mp. 4352 Display 28 follow up study.mp. 13370 Display 29 prospective studies.mp. 7126 Display 30 prospective study.mp. 69402 Display a 31 (controls or prospective$ or volunteer$).mp. 2443904 Display 32 or/22-31 2504125 Display { 33 32 not 9 2459867 ' Display,-.; 34 33 not (10 or 21) 1976531 Display 35 10 or 21 or 34 3161286 Displ ay 36 arterial hypertension.mp. 7226 Display 37 hypertension.mp. 170168 Display 38 high blood pressure.mp. 4495 Display 39 elevated blood pressure.mp. 2162 Displ ay 40 hypertensive patients.mp. 12971 Display 41 or/36-40 172551 Display 42 target level.mp. 301 Display 43 target blood pressure.mp. 219 Display 44 target systolic blood pressure.mp. 8 Display j 45 target diastolic blood pressure.mp. 33 Display 46 intensive treatment.mp. 1584 Display 47 intensive blood pressure treatment.mp. 1 Display 48 intensive antihypertensive treatment.mp. 13 Display 49 intensive control.mp. 67 Display 50 intensive blood pressure control.mp. 15 Display 51 tight control.mp. 513 Display j 52 tight blood pressure control.mp. 54 Display 53 strict control.mp. 537 , Display j 95 54 strict blood pressure control.mp. 49 Display 55 or/42-54 3346 Display 56 41 and 55 552 Display 57 35 and 56 486 Display E B M Reviews - Cochrane Central Register of Controlled Trials < 1st Quarter 2004> # 1 Search History Results Display i ! randomized controlled trial.pt. 173107 Display r .2 • randomized controlled trials.mp. ; 6317 Display 3 randomized controlled trial.mp. 18697 Display 4 controlled clinical trial.pt. 62088 Display '• • 5 : controlled clinical trials.mp. 1 814 Display ] 6 controlled clinical trial.mp. 3746 Display ; 7 random allocation.mp. [ 21950 Display 8 or/1-7 248997 Display 9 exp animal/ 4277 Display 10 8 not 9 244720 Display 11 clinical trial.pt. 231194 Display ! ! 12 clinical trials.mp. 40288 Display clinical trial.mp. j 35083 Display i [ 14 \ exp clinical trials/ ! 37049 Display | 15 (clin$ adj25 trial$).mp. \ 78337 Display 16 random$ .mp. 178980 Display 1 1 7 • exp research design/ j 75179 Display 18 research design.mp. 2540 Display J f 19 ; or/11-18 i 291554 . Display 20 ; 19 not 9 287277 • Display 21 20 not 10 42712 Display i f 22 : comparative studies.mp. 551 •VS.-?™ , Display-j 23 [ comparative study.mp. 93031 Display F 2 4 ; 1 exp evaluation studies/ 47406 Display 97 25 j evaluation studies.mp. 3971 r'Displ ay 1 26 s evaluation study.mp. 171 ' Display j 1 27 ; follow up studies.mp. 19791 i Display 28 follow up study.mp. 1248 Display ] 29 prospective studies.mp. 34761 Display ] j 30 prospective study.mp. 7268 [ Display^] ! (controls or prospective$ or volunteer$).mp. ; 186603 Display i 32 or/22-31 ; 250734 Display 33 32 not 9 247592 Display j | 34 ; 33 not (10 or 21) 27289 Displ ay 1 35 10 or 21 or 34 314721 Display ... «j 36 arterial hypertension.mp. 810 I Display * . 37 hypertension.mp. 16696 FDisplay, j j | 38 high blood pressure.mp. 378 Display I | 39 elevated blood pressure.mp. ; 170 Display | 40 ; hypertensive patients.mp. 3425 Display ! ! | 41 , or/36-40 ; 17096 Display _ | M [ 42 ; target level.mp. \ 57 Display [ 43 : target blood pressure.mp. ; 70 Display ! j 44 • target systolic blood pressure.mp. 3 Display j i target diastolic blood pressure.mp. _ • Display 46 intensive treatment.mp. 241 Display " j 47 i intensive blood pressure treatment.mp. -, 48 ; intensive antihypertensive treatment.mp. 3 Displ ay j 49 ; intensive control.mp. j 13 Display --I | 50 : intensive blood pressure control.mp. 7 Displ ay 1 51 ; tight control.mp. 20 Display H 52 ; tight blood pressure control.mp. 12 Di splay i 53 : strict control.mp. 39 Display I 5 4 strict blood pressure control.mp. 11 Display 55 | or/42-54 480 Display 1 r 5 6 • 41 and 55 138 Display 1 35 and 56 | 137 Display i 98 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0091643/manifest

Comment

Related Items