Investigation and management of resistant hypertension: British and Irish Hypertension Society position statement

Use of Validated Equipment

It is essential that clinically validated BP devices are used, noting that monitors over 4 years old may be inaccurate [32]. Please refer to the British and Irish Hypertension Society’s website, www.bihs.org.uk, for the current list of validated blood pressure monitoring devices.

Cuff size

It is important to use a cuff that is specifically made for the BP monitor and is appropriately sized for the individual [33]. The bladder length should be 75–100% of the arm circumference and the width 35–50% of the arm circumference. Poor cuff choice may lead to inaccurate diagnosis. For example, an under-sized cuff may require higher pressure to occlude blood flow in the underlying artery, leading to a false diagnosis of RH. If there is doubt over which size cuff to use, err on the side of a larger cuff as the measurement error will be significantly less than if the cuff is too small.

Cuff placement

The centre of the cuff bladder should be placed directly on the skin over the brachial artery at the level of the heart with the lower cuff-edge 2 cm above the antecubital fossa. Cuffs placed over clothing may overestimate BP, leading to a false diagnosis of RH.

Individual factors

BP can be affected by many factors including ambient noise, low temperature, stress, full bladder, recent exercise and recent (within 30 min) ingestion of caffeine or tobacco. We recommend clinic BP is measured after 3–5 min rest, with the subject seated silently with the legs unfolded, back supported and arm resting at heart height. Clothing over the upper arm should be removed and not rolled up to avoid a tourniquet effect. Talking should be avoided during and between measurements. BP should be measured at least 3 times at one-minute intervals. The average of the last 2 measurements should be recorded. In people where there is a significant difference in BP between arms, the arm with the higher measurements should be used.

Abnormal heart rhythms

Semi-automatic BP measurement devices are unreliable among people with abnormal heart rhythms including bradycardia, frequent extrasystoles and atrial fibrillation. In such individuals, manual BP measurement using auscultatory techniques should be used.

Obesity

Excess body fat and visceral adiposity are well established contributing factors in the development of hypertension and can also enhance salt sensitivity and promote vascular dysfunction and activation of the autonomic nervous system [40, 41]. There is a strong linear correlation between increasing Body Mass Index (BMI) and rise in BP [42–44]. Obesity is independently associated with RH [45] and the prevalence in this population is over 50% [21, 46]. In the National Health and Nutrition Examination Survey a BMI > 30 kg/m² approximately doubled the risk of RH among hypertensive individuals [21]. Obesity also clusters with other risk factors and clinical features of RH including obstructive sleep apnoea (OSA)—another independent cause of raised BP [47]. Weight loss strategies most commonly combine caloric, fat or carbohydrate restriction with increased exercise. A 6–8% reduction in body weight is associated with a 5/4 mmHg reduction in SBP/ DBP, but unfortunately is difficult to maintain long-term [48–52]. The latter is unsurprising given the biological basis for obesity [53]. Indeed, obesity management has been revolutionised by GLP-1 agonists which stimulate insulin secretion, down-regulate appetite via central mechanisms and delay gastric emptying. These drugs were initially developed to manage diabetes, but their effectiveness was accompanied by significant weight loss and reduction in major adverse cardiovascular events leading to subsequent studies in non-diabetic patients [54, 55]. Weight reduction obtained by the combination of a GLP-1 agonist and glucose-dependent insulinotropic polypeptide (GIP) agonists has been compared to that produced by bariatric surgery [56]. These drugs are currently licensed in the UK for the treatment of obesity, but global supply shortages are currently limiting their use in clinical practice [57].

Dietary salt

Studies have clearly linked dietary salt with BP level although there is large inter-individual variation [58]. Several studies have suggested a high prevalence of excessive salt intake among people with RH [59–61]. In addition, although “salt sensitivity” is a continuous phenotype [62], the magnitude of BP rise in people with RH that can be causally attributed to dietary salt intake is considerably larger than that observed in the general hypertensive population [63]. Reducing dietary salt intake to less than 5–6 g/day per day should be encouraged in all people with hypertension. A recent meta-analysis among people on antihypertensive therapy showed a reduction in salt of 3 grams/day resulted in a decrease in BP of ≈3.5/2 mmHg [64]. This study also highlighted that salt restriction was particularly effective at reducing BP among those treated with renin angiotensin aldosterone system (RAAS) inhibitors. Evidence from randomised controlled trials (RCTs) show the beneficial effects of a low salt diet compared to a high salt diet in a small number of people with RH [63] as well as in people with CKD with prevalent RH [65]. It is important that medication reviews are conducted to ensure people are not on concurrent medication with a high salt content, to avoid iatrogenic hypertension [66, 67].

Dietary potassium

The landmark SSaSS trial (Salt Substitute and Stroke Study) demonstrated salt substitutes with reduced sodium and increased potassium were associated with a reduction in both BP and major adverse cardiovascular events [68]. A recent meta-analysis provides support for a population goal of potassium intake of 90 mmol per day [69]. In most trials, potassium supplementation was achieved by administration of potassium chloride pills, but the reduction in both BP and major adverse cardiovascular events was similar when dietary modifications were used [68–70]. Because potassium-rich diets tend to be heart-healthy, they are preferred over the use of pills for potassium supplementation. However, the use of potassium-enriched salt substitutes should be encouraged for home cooking.

Diet

The most notable among the individual strategies are the “Dietary Approaches to Stop Hypertension” (DASH) diet, which emphasises lean proteins, whole grains, vegetables and fruits, and low-fat dairy products [71, 72]. The DASH diet reduces BP in people with hypertension (SBP reduction of 11.6 mmHg) but also in normotension (SBP reduction of 5.3 mmHg) [71, 73]. There are however no RCTs with the DASH diet as an intervention alone specifically in RH.

Physical inactivity

Both reduced physical activity and lower physical fitness are independent risk factors for hypertension [74]. Although the data to show a correlation between physical inactivity and RH are limited, the prevalence of self-reported physical inactivity is high in people with RH (~40%) [45]. There is good evidence supporting the benefits of exercise therapy for hypertension [75], and several studies have reported benefits of exercise therapy among people with RH [76–83].

Alcohol

A causal association between alcohol consumption and increased BP has been established by RCTs and Mendelian randomisation studies [84, 85]. Around a third of people with RH have a history of excessive alcohol intake [86], and people who consume excess alcohol are more likely to develop RH during treatment for hypertension [38].

Steps 1–4: Triple Therapy

Steps 1–4 of the pharmacological management of RH is to ensure that the A + C + D regimen is optimised, both in terms of choice of agents and dose (Fig. 3). A practical guide to therapeutic options is available in the BIHS statement “Adult hypertension referral pathway and therapeutic management: British and Irish Hypertension Society position statement” [114].

Fig. 3 British and Irish Hypertension Society Adult Hypertension Pathway Therapeutic Management. Reproduced from Lewis et.al. J Hum Hypertens. 2024;38 [1]: 3-7 [114] under Creative Commons licence CC BY 4.0: http://creativecommons.org/licences/by/4.0/.

Footnotes to Fig. 3: See Lewis et al. J Hum Hypertens. 2024;38 [1]: 3-7 for footnote explanations [114].

In general, long-acting drugs are to be preferred, as they provide smoother BP control and are less affected by missed doses, as are agents with robust efficacy data. Under-dosing is particularly common with ACE-I/ARB, as very low doses are available for initial treatment in people with heart failure. Switching from thiazide to thiazide-like diuretic and from indapamide to chlorthalidone may be useful in people with RH and renal impairment, if eGFR > 15 mL/min/1.73 m², as chlortalidone is the only thiazide-like diuretic to demonstrate efficacy in people with advanced CKD [115]. In subjects with CKD, particularly in cases of signs or symptoms suggestive of fluid retention, a long-acting loop diuretic such as torsemide [116] may be considered in addition or instead of a conventional thiazide-like agent (Fig. 4).

Fig. 4 Therapeutic management of resistant hypertension.

BP blood pressure, HT hypertension, CKD chronic kidney disease, HF heart failure, T2DM type 2 diabetes mellitus, DOT directly observed therapy.

In people unable to tolerate standard doses of amlodipine due to ankle oedema, switching to a more lipophilic agent - such as lercanidipine or lacidipine may be helpful [117]. Alternatively, non-dihydropyridines CCB such as long-acting formulations of verapamil or diltiazem (e.g., diltiazem LA 200/300 mg) may be used, assuming there are no cardiac contra-indications.

In general, SPCs are preferred as they improve adherence and lead to better BP control, compared with single-drug equivalent combinations [29]. However, there is limited availability of suitable SPCs in the UK, as combinations generally do not contain standard doses or drugs. For example, the only dual SPC available in the UK containing indapamide has only 5 mg of perindopril, and the only triple SPC contains olmesartan, amlodipine and hydrochlorothiazide. Although hydrochlorothiazide is commonly used as monotherapy in the USA and is a component in several SPCs in the UK, it is not a first-line NICE-recommended diuretic. While often considered to be less efficacious than chlorthalidone (dose-dependent results), hydrochlorothiazide was not inferior to chlorthalidone in preventing major CV outcomes in a large pragmatic trial (albeit with some design limitations) that randomised more than 13,000 people in the Department of Veterans Affairs Health System [118]. Similar findings were reported from two other large cohort studies [119, 120]. Low-dose thiazides have no placebo-controlled RCT evidence of mortality or morbidity benefits and concerns have been raised about the risk of skin cancer with long-term use of hydrochlorothiazide, and the Medicines and Healthcare products Regulatory Agency (MHRA) has issued guidance on its use in the UK [121]. People need to be warned of the cancer risk and the need to avoid sun exposure, and it should not be used in people with a history of skin cancer [121, 122].

Step 5: Consider initiating Spironolactone

If BP remains elevated despite optimal triple therapy, a 4th line drug should be added. There are several different drug classes available, including alpha receptor antagonists, beta-blockers and mineralocorticoid receptor antagonists (MRA). The PATHWAY-2 RCT compared these three classes head-to-head and reported spironolactone was the most effective agent in lowering BP [123]. Spironolactone was well tolerated in the ASCOT study as a third- or fourth-line agent, with a low rate of reported side-effects [124]. These data highlight the importance of salt and water retention in RH, and support the view that aldosterone excess, that does not fulfil strict criteria for classical PA, may be responsible for resistance to commonly used antihypertensive medications [125].

Spironolactone is the only MRA licensed in the UK for RH and should be used as the 4th line drug of choice. [123, 126–128] It should be initiated at 12.5–25 mg od (if eGFR > 30 mL/min/1.73 m² and K⁺ < 4.5 mmol/L) and can be up-titrated to 50 mg based on clinical response. The use of spironolactone may be limited by poor tolerability in a small number of people, such as significant hyperkalaemia, gynaecomastia or erectile dysfunction in men, and menstrual irregularities and breast tenderness in women. In such situations amiloride (5–20 mg) can be used as an alternative. In a sub-study of PATHWAY-2, people with RH were given amiloride 10–20 mg, which proved as effective in lowering BP as spironolactone, and was well tolerated [129]. Amiloride may be particularly effective in individuals with low levels of renin and aldosterone, and in black people in whom a Liddle-like phenotype appears relative common (~10–20%) [130, 131]. Eplerenone may be used as an off-label alternative in people with adverse effects due to spironolactone or amiloride. Although eplerenone is more selective (thus potentially causing fewer side effects), it is also a less potent antagonist of the mineralocorticoid receptor, and shorter acting, compared with spironolactone. In general, eplerenone must be dosed twice as high as spironolactone for therapeutic equivalence [132].

If the addition of spironolactone or equivalent is not tolerated or contra-indicated, e.g., clinically significant hyperkalaemia, then consider referral to a hypertension specialist for advice on alternative agents.

Step 6: Add in other drugs

There is a wide choice of additional agents available, albeit with limited RCT evidence to support widespread use, including but not limited to

• α-adreneceptor antagonists e.g., Doxazosin XL 4–8 mg.
• β-adreneceptor antagonists e.g., bisoprolol or nebivolol, both 2.5–5 mg OD.
• Centrally acting drugs e.g., methyldopa, clonidine or moxonidine
• Direct vasodilators e.g., hydralazine or minoxidil.

The choice of agent should be carefully considered for each patient, taking into consideration patient choice, previous adverse experiences and the clinical situation.

Adrenoceptor antagonists are perhaps better tolerated in general than the other classes of drug, and PATHWAY-2 provided evidence of efficacy for Doxazosin XL [123]. Therefore, they are a sensible initial choice.

Use of non-modified release doxazosin is not recommended due to higher peak concentration [133], which may produce rapid reductions in BP and is less well tolerated.

β-blockers may be considered in subjects with a heart rate ≥60 bpm with indications for their use e.g., migraine. If they are contra-indicated or poorly tolerated, α-blockers are an alternative.

Centrally acting drugs are effective antihypertensives but are associated with significant side effects such as depression, dry mouth, somnolence and insomnia. Although clonidine, a centrally acting α₂ agonist showed a similar BP-lowering efficacy to spironolactone in RH, it appears less well tolerated [134], and requires more careful and sustained dose titration. Moreover, there is a risk of rebound hypertension during periods of nonadherence, and it must be withdrawn slowly to avoid this. Transdermal formulations could be preferred but are currently not licensed in the UK. Methyldopa is now rarely used outside pregnancy, due to its poor tolerability and challenges with dose titration. Like clonidine, it should be withdrawn slowly.

Hydralazine and minoxidil are direct acting vasodilators and can cause profound hypotension. Both are licensed in the UK for severe hypertension. Chronic use leads to fluid retention and tachycardia – so-called pseudo-tolerance – which reduces their effectiveness. Hence, they are usually given with β -blockers and loop diuretics to counteract this. Hydralazine can produce a lupus-like syndrome particularly in those genetically defined as slow acetylators, and minoxidil is associated with excess hair growth and, rarely, pericardial effusion.