Skip to main content



Clinical courses


Clinical research courses

Raj Kishor
Avigna Clinical Research Institute
Bangalore, India

The pharmacologic treatment of hypertension (HT) has been extensively studied by clinical trials. These studies have provided definitive evidence of treatment benefit and the weight and consistency of the clinical evidence has lead to uniformity in many aspects of treatment recommendations worldwide. The international guidelines and current clinical and biostatistical practices were reviewed for relevant clinical, design, end-point assessments and regulatory issues. The results are grouped mainly into ethical, protocol and assessment issues. Ethical issues arise as placebo-controlled trials (PCTs) for HT-lowering agents in patients with moderate to severe HT are undertaken. Patients with organ damage due to HT should not be included in long-term PCT. Active-control trials, however, are suitable for all randomized subsets of patients, including men and women, and different ethnic and age groups. Severity subgroups must be studied separately with consideration to specific study design. Safety studies must be very vigilant on hypotension, orthostatic hypotension and effects on heart. In dose-response studies, at least three doses in addition to placebo should be used to well characterize the benefits and side-effects. Mortality and morbidity outcome studies are not required in anti-HT trials except when significant mortality and cardiovascular morbidity are suspected. Generally, changes in both systolic and diastolic blood pressures (BP) at the end of the dosing interval from the baseline are compared between the active and the control arms as the primary endpoint of anti-HT effect.


The world wide increment of hypertension (HT) and cardiovascular diseases (CVD) is becoming heavier than ever before with each passing year. It is estimated that by 2025, up to 1.58 billion adults worldwide will suffer from some complications of or from HT. [1] that makes one out of each three adults, on an average, will develop clinical HT or its co-morbidities or both. Currently, the prevalence of HT varies around the world, with the lowest prevalence in rural India (3.4% in men and 6.8% in women)[2] and the highest prevalence in Poland (68.9% in men and 72.5% in women).[2] However, in fact, the low-prevalence rates, e.g. as those cited for India, do not necessarily mean a really low occurrence of the disease in this population. Even in those who are diagnosed with HT, treatment is frequently inadequate. In any case, regardless of the prevalence rate, large or small, HT and related diseases must be intervened for prevention, diagnosis and control.There are numerous trials in hypertension many of which have focused on cardiovascular(CV) outcomes (death, non–fatal Myocardial Infarction, stroke, congestive heart failure).These trials have influenced our clinical practice in terms of setting out guidelines in the treatment of Hypertension.


In the late 1960s to early 1970s the trials were carried out in rather select populations (both high- and low-risk elderly patients), and were mainly interested in efficacy measurement of the anti-HT drugs in lowering BP relative to placebo or no treatment. The trials of the 1980s focused on middle-aged hypertensive individuals and later on the elderly. These trials mainly investigated the efficacy of β-blockers and diuretics in reducing the systolic BP in these patients. These early efficacy trials did not address the effect of anti-HT agents in controlling serious co morbidities like MI, CHD and CHF. Their compliance with Good Clinical Practices (GCP) was also rather poor. [3]

During the 90s, large randomized trials in a much broader patient population came in vogue and a great portion of these trials elucidated whether and whom to treat with different classes of anti-HT drugs. Continuing to date, these trials of angiotensin converting enzyme inhibitors (ACEIs), β-blockers, angiotensin receptor blockers (ARBs) and calcium channel blockers (CCBs) investigated the exact end-point goals, such as reduction of a certain percent of risk of MIs in Grade II HT patients over 1 year of treatment. The trials evaluated, often, the superiority of the end-point reduction by one agent over another as opposed to the overall efficacy of one agent or the other. Reduction in the relative risk of mortality from the primary and secondary outcome measures is one of the main objectives of the RCTs over the last 5–7 years. [3, 4]The reason for the outcome measurement shifting from lowering of BP to other parameters is that for some drugs, the BP lowering is an inadequate marker (surrogate) of health benefits in HT. Anti-HT drugs can have other important actions that may alter the benefit of lowering of BP. For example, many anti-HT drugs have shown consistent beneficial effects on long-term mortality and morbidity, most clearly on stroke and less consistently on cardiovascular events, such as, low- and high-dose diuretics, reserpine and β-blockers, usually as part of combination therapy. [5]

The trial patient population:
For participation in RCTs for HT or its co morbidities, any patient with a BP >120/80 mmHg, especially with one additional risk factor such as body mass index >25, is at risk of developing clinical HT and may qualify for the trial. For example, BP values between 130 and 139/85 and 89 mmHg are associated with a more than two-fold increase in the relative risk from CVD as compared with those with a BP of 120/80 mmHg or below.[3] Patients who are of 55 years of age or more and are also obese or diabetic are particularly risk-prone.

Currently, the patient population studied with a new anti-HT includes a broad range of patients with HT and its co morbidities. For mild to moderate HT, however, only BP can be studied in a CT by measuring both DBP and SBP over the study period. More severe HT is usually studied with relevant concomitant illness, e.g. CHD and DM. Care should be taken that the drugs they need would not interfere with the observations of the effects of the study drug. For example, for patients with CHF, standard treatment requires use of one to several agents (ACE inhibitors) affecting BP, which could have pharmacological actions similar to those of the study drug.[5]

Grading of HT together with target organ damage (TOD) secondary to HT needs to be established accurately. Patients, e.g. with BP >160/110 mmHg and DM, cannot be included in PCTs. Such patients, however, can be included in active-controlled trials with proper safety monitoring. Patients from relevant demographic subsets should be studied, including both men and women, racial/ethnic groups pertinent to the region and both young and older patients. The very old or “fragile elderly,” i.e Patients >75 years old, should be included.

In general, all population subsets should be included in the same studies rather than conducting studies in subgroups. This facilitates comparisons across subsets in the same environment. An exception would be severity of the subgroups, where study designs could be different for different severities. Patients with secondary HT, isolated systolic HT, HT during pregnancy and children with HT should be studied separately if specific indications for use in those populations are being sought. [5]

RCT’S of antihypertensive agents:

All recent trials, since the 1990s, for the assessment of efficacy and safety of anti-HT drugs have been designed and conducted as randomized, blinded trials.[6] Such trials are not only free of experimental bias but they are also balanced in all important aspects of the study and differ only in the intervention that the experimental and the control groups receive. As discussed further in the following section, many big trials, such as INVEST (The International Verapamil –Trandolapril study hypertension) and ALLHAT(Antihypertensive and lipid lowering to prevent heart attack trial), are designed as prospective, randomized, open, blinded-end point evaluation (PROBE) investigations in thousands of patients in multiple countries. In some studies, there is also an emphasis on the determination of the reduction of mortality and/or CV morbidity by the experimental treatments rather than measuring just the BP-lowering effects.[7] Well-designed and well-conducted RCTs have been able to estimate such complex end-points with a great deal of success.

Study design and Randomization:
The long-term HT trials are designed today as PROBE(Prospective Randomised,open,blinded end point) studies. Such studies are aimed at comparing a treatment regimen of newer anti-HT drugs (e.g, a CCB) with a traditional regimen (β-blocker and/or a diuretic), like prevention of CHD. The study basically consists of two arms, i.e. the control and the experimental arms, in which appropriate number of patients are entered following a randomization scheme. Because the control arm in such studies consists of receiving an active drug, which is often one of the standard first-line treatments, the trial is often dubbed as an “active control trial,” as mentioned above several times. A schematic representation of a two-arm randomized trial is presented in Figure (1)below.

The total number of patients (sample size) and those in each arm are calculated carefully such that a clinically meaningful effect size can be differentiated between the average outcome measurements for the two arms with adequate statistical power (usually 80%) and significance (usually two-sided 5%). Both the power and the level of significance are prospectively defined and finalized in the detailed protocol before the trial begins. The final sample size includes considerations of drop out patients and all interim analyses (IA)7 Schematic representation of a two-arm randomized clinical trial of new and existing anti hypertension (anti-HT) treatments in HT patients with one or more additional risk factors.

The RCTs for HT and co morbidities are usually large (few thousand patients) multi centric studies.[6]Open label refers to a non concealment of both the active control and the experimental drugs to the patients and investigators, in that they can figure out the difference in physical and organoleptic properties between the two. Sometimes, there could also be a difference in the route of the two administrations.

The investigator or the expert who measures the endpoint, however, is blinded to the randomization codes and allocations to all patients, such that no bias is introduced in the assessment of the endpoint. If dosage titration for the experimental arm is required, the same principle of endpoint blinding should be applied.[7]

There are three basic ways to generate a randomization scheme for an RCT.
Simple randomization
2. Block randomization
3. Stratified blocks

Simple randomization: The simple randomization, which is equivalent to tossing a coin for each subject that, enters a trial. The heads get the experimental treatment while the tails receive the placebo. A computerized or tabulated random number generator is generally used. It is simple and easy to implement and treatment assignment is completely unpredictable.

Block randomization: It is very popular and balanced within each block. For a trial of n treatments, the total number patients are divided into m blocks of size 2n. Then, each of the m blocks is randomized such that n patients are allocated to each of the treatments. One can then choose the blocks randomly. The INVEST[15] study followed this scheme of block randomization.

Stratified blocks: , a third approach to randomization involves “stratified blocks.” Because a trial may not be considered valid if it is not well balanced across the prognostic factors, stratification of patients is done to produce comparable groups with regard to certain characteristics (e.g., gender, age, race, disease severity). This approach produces valid statistical tests in all stratified subgroups (e.g., high-risk subgroups in the ALLHAT trial).[13,14

Whatever the mode of randomization is, it is ensured that the pattern of assignment of control or experimental drug within a group of patients cannot be guessed at any point. It is recommended that the statistician who generated the randomization codes does not get involved in the IA or the final analysis of the experimental data. Other study designs can be used in HT trials as long as they are scientifically valid and manageable.[10]

Usually, studies are designed for observation and analysis of the primary outcome on which the sample size calculation is also based. Secondary outcomes, however, can also be validly analyzed if the primary outcome difference is not statistically significant provided that they were declared a prior and are clinically important. Another condition for the valid use of secondary outcomes in the efficacy or endpoint estimation is that the method to capture outcomes was the same in each treatment group and the data are unbiased (randomized). In addition, if the outcomes for secondary endpoints such as heart failure (HF) and CVD are still compelling even after considering the number of comparisons made, then the conclusion based on these outcomes is valid.

Assessment /Analysis of the study:

The primary basis of assessment of efficacy of antihypertensive drugs is the effect of the drug on systolic and diastolic blood pressures. In the past the primary endpoint of most studies was diastolic blood pressure. Although all drugs to date have reduced both systolic and diastolic blood pressures, the recognition of isolated or predominant systolic hypertension as a significant and remediable risk factor demands explicit evaluation of the effect of a drug on systolic blood pressure. Many clinical trials of many interventions (including low and high dose diuretics, reserpine, and beta-blockers, usually as part of combination therapy) have shown consistent beneficial effects on long-term mortality and morbidity, most clearly on stroke and less consistently on cardiovascular events. Whether some drugs or combinations have better effects than others on overall outcomes or on particular outcomes is not yet known. Formal mortality and morbidity outcome studies are not ordinarily required for approval of antihypertensive drugs and the kind of active control mortality and morbidity studies that would be convincing is not well defined. Results of a large number of on-going outcome studies could affect this policy and modify requirements. It should be noted that, even if an antihypertensive effect has been proven, a significant concern about a detrimental effect on mortality and/or cardiovascular morbidity might lead to a need for outcome studies.[11]


2.1 Studies to assess antihypertensive effect:  The primary endpoint of studies to assess antihypertensive effect is the absolute change at the end of the dosing interval (trough) from drug-free baseline blood pressure compared to the change in the control group. As a secondary endpoint effects can also be assessed with respect to pre-defined response criteria. In general, the effect on blood pressure at the end of the study is the primary endpoint, but the time course of the onset of the effect is also of interest; this can be defined by examining trough response each week or every two weeks in some studies. [11] The effect on blood pressure and the relationship of the response to dose should be characterized in short-term studies (4-12 weeks), whose short duration allows for use of a placebo control. Long-term (six months or more) studies should also be carried out to demonstrate maintenance of efficacy and to look for withdrawal effects. Because blood pressure readings are subject to systematic error (bias), because spontaneous changes in blood pressure can be large, and because the effect of active drugs is often small  studies conducted in a blinded fashion and with placebo controls are essential . In general, short-term studies should be placebo-controlled. Dose-response studies and studies using an active control drug as well as placebo are strongly encouraged. Controlled randomized short-term studies of various designs can be useful in demonstrating effectiveness, for example:
A) Single fixed-dose vs. placebo
B) Optional titration (based on response) vs. placebo
C) Forced titration vs. Placebo
D) Fixed-dose, dose-response vs. placebo (can use forced titration to reach the randomly assigned fixed maintenance dose)
E) Any of the above designs with an active control drug

2.2 Pharmacodynamic studies: The Pharmacodynamic properties of antihypertensive drugs should be characterized. Studies should be performed to evaluate such properties as hemodynamic effects, renal effects, and neurohumoral effects.[13] In general, it is useful to characterize the magnitude, dose-response, and time-course of these effects. These studies should usually be placebo-controlled.

2.3 Dose response relationship: The dose-response (D/R) relationship for favorable (blood pressure) and unfavorable effects of anti-hypertensive drugs should be well-characterized through randomized fixed-dose, D/R studies. In these studies, a greater number of dose groups will provide a better D/R assessment. If possible, at least three doses (in addition to placebo) should be used. Although trials usually use a randomized parallel fixed-dose, D/R design (See ICH E-4), some studies could utilize a placebo-controlled titration design, appropriately analyzed, to narrow the range of doses to be studied in fixed-dose studies, and to characterize individual D/R relationships. Either or both of these designs can provide evidence for anti-hypertensive efficacy. These D/R studies should characterize critical parts of the D/R curve, allowing identification of the lowest dose with some useful effect, a dose on the steep part of the D/R curve, and a dose beyond which further effects are absent or small (maximum useful dose). If there is a positive D/R slope, a D/R study can show effectiveness even without a placebo, but it may provide littleinformation on the value of the lower doses (which may then need further study). A D/R study may be uninformative if all doses show equal effects.[14]

2.4 Comparison with standard therapy: The need for studies comparing the new drug with current standard therapy varies among regions but comparative studies are of interest in all regions. In order to assess antihypertensive efficacy, these trials need to document assay sensitivity through a placebo arm or a terminal placebo-controlled randomized withdrawal phase. For the short-term studies, a 3-arm design(test drug, active control drug, and placebo) may be particularly helpful as a study that not only supports efficacy but also makes a comparison with standard therapy. If the effect size (vs.placebo) is unusually small, it can be helpful to know whether this was the result of the study population, or other study features (test and control drugs both have small effect vs. placebo), or the drug (active control drug has a larger effect than test drug compared to placebo). For longer-term  comparative studies, where concurrent placebo-control is not possible.


ICH E-1 suggests that a database of about 1500 patients (300-600 for 6 months, 100 for 1 year) is usually sufficient for chronically administered drugs, but as suggested in that guideline; this may be too small for the very long, very wide exposure in an asymptomatic population intended for antihypertensive drugs. [15] In addition to the usual safety assessments, attention should be paid to excessive fall in blood pressure (hypotension), especially on standing (orthostatic hypotension), and rebound phenomena. Depending on the particular drug and other observations, studies of effects on heart rhythm or cardiac conduction, coronary steal effects, effects on risk factors for cardiovascular disease (e.g., blood glucose, lipids), and effects on target organ damage are of interest.

As antihypertensive treatments are often used in combination, it is important to study the efficacy and safety of the new drug in this situation. Information on combination use can be obtained in formal factorial studies and in combined use in the course of long- and short-term clinical studies. Specific requirements regarding clinical co-administration studies may vary among regions. Studies in patients with inadequate blood pressure control on other agents will provide information on the effect of the test drug when added to other agents (add on studies). Studies in which additional drugs are added to the test drug to achieve an adequate response may also be useful.

There are two approaches to the combination studies needed to obtain safety and efficacy data to support fixed combination products. It is essential to consult with regional regulatory authorities regarding the specific data needed to support the specific indications.

5.1 Factorial study:  In a factorial study, placebo, and one or more doses of the test drug T and another drug D are studied alone and in the combination in a short-term randomized controlled trial. Such a trial can be used to show that the combination has a greater effect than either drug alone. Most informative for identifying the appropriate dosages for a fixed combination is a factorial D/R study in which several doses of each drug, e.g., a test drug T and another drug D, and their combinations are compared, as shown below.

This design reveals D/R relationships for the test drug and other drug alone and in combination with each dose of the other treatment and may support one or more fixed combinations. These studies reveal "response surface" relationships that use data from all groups. It may be necessary to study separately the low doses and low dose combinations to establish their specific usefulness.[16]

5.2 Studies in non-responders to each drug
The safety and efficacy of combinations can also be assessed by examining the effect of the combination in patients failing to respond to both of the single drugs, e.g., diastolic blood pressure > 90 mmHg on that component. In some cases, regional authorities only require trials in patients failing to respond to one of the components.

This article deals with control and design principles that guide the conduct and conclusion of a meaningful HT trial. The international guidelines and current clinical and biostatistical practices were reviewed for relevant clinical, design, end-point assessments and regulatory issues.

The results are grouped mainly into ethical, protocol and assessment issues. Each trial is as meaningful as the number of scientific questions it answers and paves the direction of future trials in that field.

Compliance to GCP opens up the data for correct interpretation. Any neglect toward this end of abiding by the GCP principles can raise a batch of critical questions, eventually rendering the entire data to be uninteresting or suspect.

Principles of study design and analysis mentioned in this article can give the trial lists an advantage of a well-designed study, providing the crucial evidence of safety and efficacy of the agent under testing.

1. Kearney PM, Whelton M, Reynolds K, Muntner P, Whelton PK, He J. Global burden of hypertension: Analysis of worldwide data. Lancet. 2005;365:217–23
2. Kearney PM, Whelton M, Reynolds K, Whelton PK, He J. Worldwide prevalence of hypertension: A systematic review. J Hypertens. 2004;22:11–9.
3. Gavras I, Gavras H. Benefits and side effects of blood pressure lowering treatment: What was wrong with doxazosin in the ALLHAT? Curr Control Trials Cardiovasc Med. 2001;2:257–9.  
4. Rahn KH. Recent intervention studies with antihypertensive drugs and their influence on guidelines.Med Klin. 2003;98:771–5.
5. Centre for Drug Evaluation and Research, USA Food and Drug Administration. Guidance for industry - E12A: Principles for Clinical Evaluation of New Antihypertensive Drugs. 2002 Available
6. Brenner BM, Cooper ME, de Zeeuw D, Keane WF, Mitch WE, Parving HH, et al. Effects of losartan on renal and cardiovascular outcomes in patients with type 2 diabetes and nephropathy.(RENAAL) N Engl J Med. 2001;345:861–9.
7. Dahlof B, Sever PS, Poulter NR, Wedel H, Beevers DG, Caulfield M, et al. Prevention of cardiovascular events with an antihypertensive regimen of amlodipine adding perindopril as required versus atenolol adding bendroflumethiazide as required, in the Anglo-Scandinavian Cardiac Outcomes Trial-Blood Pressure Lowering Arm (ASCOT-BPLA): A multicentre randomised controlled trial. Lancet.2005;366:895–906.  
8. Rubin LJ, Badesch DB, Barst RJ, Galie N, Black CM, Keogh A, Pulido T, Frost A, Roux S, Leconte I, et al. Bosentan therapy for pulmonary arterial hypertension. N Engl J Med2002;346:896–903.  
9. Olschewski H, Simonneau G, Galie N, Higenbottam T, Naeije R, Rubin LJ, Nikkho S, Speich R, Hoeper MM, Behr J, et al. Inhaled iloprost for severe pulmonary hypertension. N Engl J Med2002;347:322–329.
10. Barst RJ, McGoon M, McLaughlin V, Tapson V, Rich S, Rubin L, Wasserman K, Oudiz R, Shapiro S, Robbins IM, et al. Beraprost therapy for pulmonary arterial hypertension. J Am Coll Cardiol 2003;41:2119–2125.
11. Barst RJ, Langleben D, Frost A, Horn EM, Oudiz R, Shapiro S, McLaughlin V, Hill N, Tapson VF, Robbins IM, et al. Sitaxsentan therapy for pulmonary arterial hypertension. Am J Respir Crit Care Med 2004;169:441–447.  
12. Galiè N, Ghofrani HA, Torbicki A, Barst RJ, Rubin LJ, Badesch D, Fleming T, Parpia T, Burgess G, Branzi A, et al. Sildenafil citrate therapy for pulmonary arterial hypertension. N Engl J Med2005;353:2148–2157.
13. Center for Drug Evaluation and Research. Medical review of application #22–081: Ambrisentan. Center for Drug Evaluation and Research; 2007.
14. McLaughlin VV, Oudiz RJ, Frost A, Tapson VF, Murali S, Channick RN, Badesch DB, Barst RJ, Hsu HH, Rubin LJ. Randomized study of adding inhaled iloprost to existing bosentan in pulmonary arterial hypertension. Am J Respir Crit Care Med 2006;174:1257–1263.  
15. Simonneau G, Rubin LJ, Galie N, Barst RJ, Fleming T, Burgess G, Collings L, Cossons N, Badesch DB. Safety and efficacy of sildenafil-epoprostenol combination therapy in patients with pulmonary arterial hypertension [abstract]. Am J Respir Crit Care Med 2007;175:A300.
16. EARLY: first functional class II population study emphasizes need to diagnose and treat PAH early [media release]. Allschwil, Switzerland: Actelion Pharmaceuticals; 2007.



Subscribe to Pharmatutor Alerts by Email