This study was recommended by NICE, as part of its 2006 guidance for the treatment of hypertension, and is urgently required to provide evidence for the treatment recommendations in patients with resistant hypertension. The study will be a randomised placebo-controlled double-blind crossover comparison of an α-blocker (α), β-blocker (β), and K+-sparing diuretic (∆).

Patients will have a BP at entry above target on ABPM or home monitoring despite supervised administration of maximum tolerated doses of A+C+D. Over 48 weeks they will then receive, in random order either placebo or two doses each of doxazosin (α), bisoprolol (β) or spironolactone (∆). Each treatment cycle will last 12 weeks, with a forced dose-doubling at 6 weeks.

The time course for the study will be similar to study one. 340 patients will provide 90% power, at α=0.01 to detect a 3 mmHg overall difference in home sBP between any one drug and placebo, with spironolactone hypothesized to be best overall. The study will be able to detect a 6 mmHg difference in sBP between each subject's best and second-best drug predicted by tertile of plasma renin, justifying routine use of the measurement in patients with resistant hypertension.

In published surveys throughout the world the majority of patients with hypertension do not achieve target blood pressure. According to most guidelines including NICE, target blood pressure is 130/80 mmHg in patients with diabetes, 140/90 mmHg in other patients. In the UK there are currently at least 3 million people with treated hypertension whose blood pressure is not controlled. A significant number of these patients will have drug resistant hypertension, defined as:

"a blood pressure that is not adequately controlled to recommended treatment targets despite treatment with maximal recommended and tolerated doses of 3 drugs according to the current BHS/NICE guidelines and treatment algorithm, i.e. (*ACE-inhibitor or *ARB or direct renin inhibitor + *CCB + Diuretic (any diuretic except spironolactone), i.e. A+C+D)".

(*where ACE-inhibitor=angiotensin converting enzyme inhibitor, ARB=angiotensin receptor blocker, CCB=calcium channel blocker)

The causes of treatment resistance are unknown, and the choice of fourth-line drug almost entirely empirical. At present there is little comparative data for available drugs. There is considerable evidence pointing to Na+ retention as a common culprit, and some data supporting additional diuretics, or alpha blockade in resistant hypertension, though mainly added to two rather than three drugs.20,29,30 A retrospective analysis of two-drug combinations in trials reported that it makes no difference what is combined with what. 31 However, this conclusion conflicts with the view that drugs for hypertension fall into two main categories, acting respectively on the renin and volume components of hypertension, and that most benefit can be derived from combining drugs from different categories.10,32

Despite the successful adoption of the BHS/NICE treatment algorithm for the treatment of hypertension, there remains substantial clinical uncertainty about the preferred clinical management of people with drug resistant hypertension. This is an important question because such patients are at the highest risk for cardiovascular events. The current guidelines acknowledge that there is currently no adequate clinical trial evidence upon which to base recommendations for the preferred 4th line drug treatment for those with resistant hypertension.

It is possible that it makes no difference what drug is added as fourth-line treatment and that the response, on average, will be the same for all classes. Alternatively, it is also possible that one class of drug will always be superior to all the others because the mechanism underpinning resistant hypertension is common to all patients. In this regard, a popular view is that resistant hypertension is invariably due to excessive sodium retention and thus "further diuretic therapy" will always be the most effective treatment. A third possibility is that resistant hypertension is a heterogeneous state and that the study of average responses in cohorts in clinical studies masks substantial individual patient differences.

With regard to this, this study will address the hypothesis that the renin status of the patient defines the response to 4th line treatment in resistant hypertension, i.e. that low renin predicts sodium retention and the best response to diuretic therapy, whereas high renin predicts a better response to drugs that suppress renin, i.e. a β-blocker.

Hypothesis and Novel Aspects of the Trial

The primary hypothesis of the study is that the commonest cause of resistant hypertension is excessive Na+-retention, and that further diuretic therapy (spironolactone used in this study) will be superior to other potential "add-on drugs" for people with inadequate blood pressure control despite treatment with optimal or highest tolerated doses of the three drug classes recommended by the BHS/NICE treatment algorithm, i.e. A+C+D.

The main secondary objective is to use plasma renin measurements to evaluate an 'α, β, ∆' rule for the selection of the 4th line drug for patients with drug resistant hypertension - where α represents α-blockade, β represents β-blockade and ∆ represents "further diuretic therapy" as cited above. The main secondary hypothesis states that plasma renin (measured on a background of 3 drugs, i.e. A+C+D), will predict the most effective 4th line drug. We propose that:

α-blockade will be the most effective 4th line drug at lowering BP in patients in the mid-tertile of plasma renin, expected to be ≥20mU/L but <100mU/L; β-blockade will be the most effective drug when renin is in the top tertile (expected to be ≥100Mu/L) as the drug blocks renin secretion; Further diuretic therapy with spironolactone will be most effective when plasma renin is in the lowest tertile (expected to be <20mU/L), indicative of excessive sodium retention.

The study will also evaluate whether the routine use of plasma renin to predict best treatment in individual patients with resistant hypertension will be more cost-effective than using further diuretic therapy indiscriminately as the preferred 4th line drug for all patients.

Finally the study will investigate whether non-invasive assessment of haemodynamic parameters indicative of sodium retention and volume status, i.e. cardiac output, peripheral resistance and bioimpedance, can be used to predict the response to each drug in the α, β, ∆ sequence.

Patients must meet ALL inclusion criteria

• M/F 18-79 years

• Patients with hypertension not controlled to target: clinic systolic BP ≥ 5 mmHg above target (i.e. ≥ 140 mmHg for non-diabetic hypertensives or ≥ 135 mmHg for diabetics), under one of the following conditions:

  1. Treatment for at least 3 months with lisinopril 20 mg (A) + amlodipine 10 mg (C) + bendroflumethiazide 2.5 mg (D) or their equivalents
  2. Patients who have received the three drugs or equivalents specified in a), and are either intolerant to one category, or tolerate only a lower dose (e.g. amlodipine 5 mg or lisinopril 10 mg)
  3. Patients receiving the three drugs or equivalents specified in a), who are receiving additional drugs for their hypertension, may be included if the investigator: 1) feels it is appropriate to stop these additional drugs at the screening visit and 2) anticipates that the BP criteria for inclusion will be met when re-checked at the baseline visit Patients may be included if the PI anticipates BP criteria for inclusion will be met at randomisation.

• Patients with a home systolic BP average of >130 mmHg or within 15mmHg of clinic BP over the 4 days prior to the baseline visit.

Exclusion;

• Inability to give informed consent;

• Participation in a clinical study involving an investigational drug or device within 4 weeks of screening;

• Secondary or accelerated hypertension;

• Type 1 diabetes;

• eGFR<45 mls/min;

• Plasma potassium outside of normal range on two successive measurements during screening;

• Pregnancy, planning to conceive, or women of child-bearing potential not taking adequate contraception

• Anticipated change of medical status during the trial - Absolute contra-indication to study drugs or previous intolerance of trial therapy;

• Sustained atrial fibrillation;

• Recent cardiovascular event requiring hospitalisation

• Suspected non-adherence to antihypertensive treatment

• Requirement for study drug for reason other than to treat hypertension, - Current therapy for cancer;

• Concurrent chronic illness, likely to preclude 52 week participation in the study;

• Clinic Systolic BP >200 mmHg or diastolic BP >120mmHg, with PI discretion to override if home BP measurements are lower

• Any concomitant condition that may adversely affect the safety/ efficacy of study drug or severely limit that patients life-span or ability to complete the study

• Treatment with any of the following medications;

  1. Oral corticosteroids within 3 months of screening. Treatment with systemic corticosteroids is also prohibited during study participation;
  2. Chronic stable use, or unstable use of NSAIDs (other than low dose aspirin) is prohibited. Chronic use defined as >3 consecutive or non-consecutive days of treatment per week. In addition intermittent use of NSAIDs is strongly discouraged; if required, must not be used for more than a total of 2 days. For those requiring analgesics; paracetamol is recommended.
  3. The use of short acting nitrates is permitted, but must not be taken within 4 hours of screening or subsequent visits
  4. The use of long acting nitrates is permitted but dose must be stable for at least 2 weeks prior to screening and randomisation;
  5. The use of sympathomimetic decongestants is permitted;but not within 1 day prior to any study visit/BP assessment;
  6. The use of theophylline is permitted but dose must be stable for 4 weeks prior to screening and throughout the study;
  7. The use of phosphodiesterase type V inhibitors is permitted; however study participants must refrain from taking these medications for at least 1 day prior to screening or any subsequent study visits;
  8. The use of alpha-blockers is not permitted, with the exception of afluzosin and tamsulosin for prostatic symptoms

• A pill count will be made at the end of the 4 week run-in period and those with adherence <70% will be excluded from randomisation