Resistant hypertension is defined as a blood pressure level above target despite treatment with three optimally dosed, best-tolerated antihypertensive drugs of different classes.1 Of the three agents, one should ideally be a diuretic, one a renin–angiotensin system blocker (either an angiotensin-converting-enzyme [ACE] inhibitor or an angiotensin-receptor blocker) and one a dihydropyridine calcium-channel blocker.2 Blood pressure controlled with four or more antihypertensive agents is also considered resistant hypertension.1
Hypertension affects about 20% of Canadian adults.3,4 The prevalence of resistant hypertension in this patient population varies from 8%–12% in population-based surveys and audits of primary care practices5–7 to 11%–21% in specialized clinics.8,9 The true prevalence is difficult to ascertain because many studies failed to exclude pseudoresistance (described later). Older age, female sex, black race, excess intake of salt or alcohol, diabetes, obesity, kidney disease and long-standing, poorly controlled hypertension are associated with resistance.1,2 The risk of cardiovascular morbidity and mortality is higher among patients with resistant hypertension than among those whose hypertension is well controlled.1,2,10 This increased risk is likely mediated by uncontrolled blood pressure,10 concomitant comorbidities (diabetes, sleep apnea, obesity) and target organ damage (renal disease, left ventricular hypertrophy and cardiovascular disease).1,2,9
In this article, we review the assessment and management of resistant hypertension, including emerging therapies. A summary of our literature search is outlined in Box 1. We found few randomized controlled trials (RCTs) and no systematic reviews to guide decision-making. Thus, we have made management recommendations based primarily on expert consensus unless otherwise specified. Most of the studies identified in our literature search examined blood pressure, a validated surrogate outcome, and did not assess morbidity or mortality as outcomes.11
Evidence for this review
We searched MEDLINE (1946 through Apr. 21, 2013) using the keywords “resistant hypertension” and “refractory hypertension.” We excluded duplicate articles and limited the search to human studies that were published in English. Of 1163 citations screened, we reviewed 187 full-text articles, including 8 randomized controlled trials and 2 consensus guideline statements. We also scanned reference lists to identify additional citations of relevance.
Which factors contribute to resistant hypertension?
Resistant hypertension results from numerous and often simultaneously acting factors (Table 1 and Box 2). Obesity (especially visceral adiposity) and obstructive sleep apnea are the two most prevalent factors, the former found in 50% of people with resistant hypertension18 and the latter in 64%–83%.19,20 Concurrent use of certain medications or substances can also lead to resistance (Box 2).
Medications and substances that can increase blood pressure13
Alcohol
Nonsteroidal anti-inflammatory drugs
Oral contraceptives
Antidepressants (monamine oxidase inhibitors, certain serotonin reuptake inhibitors and serotonin–norepinephrine reuptake inhibitors)
Stimulants (amphetamines and cocaine), sympathomimetics and decongestants
Corticosteroids and anabolic steroids
Erythropoetin and analogues
Natural licorice
Herbal products (e.g., ma huang [ephedra] and bitter orange)
Chemotherapeutic agents (e.g., tyrosine kinase inhibitors and vascular endothelial growth factor inhibitors)
Primary aldosteronism, resulting from excess unilateral or bilateral adrenal secretion of aldosterone, is present in 11%–20% of patients with resistant hypertension referred to specialty clinics and is characterized by suppressed renin levels.1,21,22 High aldosterone levels are also found in many other conditions, such as obesity, sleep apnea and renal artery stenosis.1,23 Elevated aldosterone levels are associated with detrimental cardiovascular physiologic effects, such as oxidative stress, endothelial dysfunction, inflammation and tissue fibrosis.24
Elevated activity of the sympathetic nervous system is also a key contributor to resistant hypertension. This excess activity is frequently present in primary (essential) hypertension.25,26 Increased sympathetic outflow leads to systemic and renal vasoconstriction, hypertrophy and proliferation of vascular smooth muscle cells, left ventricular hypertrophy, endothelial dysfunction, insulin resistance, systemic inflammation, oxidative stress, and sodium and water retention.27
How should patients with resistant hypertension be evaluated?
The exclusion of pseudoresistance is the first step in the evaluation of a patient with resistant hypertension (Figure 1). White-coat effect (present to some extent in up to 40% of patients with apparently resistant hypertension18), inaccurate measurement techniques, nonadherence to treatment and a suboptimal medication regimen are common contributors.1 Use of 24-hour ambulatory monitoring or self-monitoring at home to document a normal out-of-office blood pressure level rules out white-coat effect.11 Manual office-based readings of blood pressure are not accurate in patients with white-coat effect; therefore, doses should be adjusted using out-of-office measurements (repeat ambulatory or home monitoring).
Nonadherence to pharmacologic and nonpharmacologic treatments should be assessed through patient interview; improved adherence reduces blood pressure.1 About 40% of patients with resistant hypertension stop treatment after one year.1 If nonadherence to pharmacologic treatment is suspected, electronic prescription records can be reviewed for confirmation. To improve adherence, consensus guidelines recommend combination products to reduce the pill burden, long-acting formulations of drugs taken once daily, self-monitoring of blood pressure and multidisciplinary patient management.1,11
After pseudoresistance is excluded and true resistant hypertension is confirmed, an assessment for secondary hypertension should be performed (Table 1 and Box 2). If feasible, medications contributing to the elevated blood pressure should be stopped (Box 2).
What modifications are most likely to be effective?
Optimizing the existing drug regimen
Selecting optimal medication combinations at the most effective dosages based on the patient’s conditions can improve blood pressure control (Figure 1). Bedtime dosing can lower nocturnal blood pressure by 5.2 mm Hg (p < 0.001) and potentially reduce mortality and cardiovascular events by 61% (p < 0.001).9,42,43 However, further trials are needed to confirm these findings in patients with hypertension and in those with resistant hypertension.11
Optimizing health behaviours
Patients with hypertension should be advised to reduce salt intake, participate in regular aerobic exercise, eliminate excessive alcohol intake, maintain a normal body weight and eat a diet based on the Dietary Approaches to Stop Hypertension (DASH) plan.11 Aside from salt reduction and aerobic exercise, the evidence base underlying these recommendations is derived largely from studies involving patients with nonresistant hypertension.44
In a four-week randomized cross-over trial, 12 patients with resistant hypertension were assigned to a low- (50 mmol/d) or high- (< 250 mmol/d) salt diet for one week, followed by a two-week washout period; the patients then followed the opposite diet for one week.45 The low-salt diet reduced systolic blood pressure by 22.7 mm Hg (95% confidence interval [CI] 11.8–33.5 mm Hg) and diastolic blood pressure by 9.1 mm Hg (95% CI 3.1–15.1 mm Hg) compared with the high-salt diet.
In a 10-week RCT, 50 sedentary patients with resistant hypertension were randomly assigned to aerobic exercise (treadmill walking three times weekly) or no exercise.41 Patients in the exercise group had significant reductions in systolic and diastolic blood pressure (6 ± 12 and 3 ± 7 mm Hg, respectively; p = 0.03 for both) compared with the control group.
Treatment of obstructive sleep apnea
Limited RCT-level data support the use of continuous positive airway pressure in patients with resistant hypertension. In an RCT involving 75 patients (65% with resistant hypertension) that compared the use of continuous positive airway pressure with no treatment, no overall between-group differences in blood pressure were noted.46 However, among patients with resistant hypertension, those who received continuous positive airway pressure had greater reductions in 24-hour ambulatory blood pressure than the controls (systolic −7.6 v. −0.6 mm Hg, p = 0.07; and diastolic −4.9 v. 0.1 mm Hg, p = 0.03).46 Although these data should be considered preliminary and will require confirmation, continuous positive airway pressure is already indicated for sleep apnea; thus, it is reasonable to recommend it to patients with concomitant resistant hypertension to help improve blood pressure control.
Add-on therapy
If blood pressure remains uncontrolled after a diuretic, a renin–angiotensin system blocker and a dihydropyridine calcium-channel blocker have been prescribed and the dosing has been optimized, the use of additional drugs can be considered (Table 2). Mineralocorticoid-receptor antagonists have been the most rigorously studied. Among the other available agents, α- and β-blockers are also commonly prescribed and have been used as add-on drugs in many large-scale clinical trials.
Selecting the add-on drug
If a compelling indication for a given class of drugs is present (Table 3), an agent from that class is chosen first.11 Otherwise, the choice can be made through renin profiling (Table 3) or empirically.
Renin profiling
Using measurements of plasma renin activity as a guide (Table 2), anti-renin agents are prescribed if renin levels are high (≥ 0.65 ng/mL per h); otherwise antivolume agents are prescribed. In an open-label RCT involving 77 patients seen in a hypertension specialty clinic (60% with resistant hypertension), medication adjustment using algorithm-guided renin profiling reduced blood pressure more than usual care (systolic −29 v. −19 mm Hg, p = 0.03; and diastolic −13 v. −11 mm Hg, p= 0.3). 47
Patient characteristics such as age and race can be used in lieu of renin measurement.48 Older patients (≥ 55 yr) and black patients often exhibit low-renin, volume-expanded hypertension and respond best to antivolume drugs.49 Younger individuals and white patients usually respond best to antirenin agents.48,49 This approach is often used to select initial therapy for treatment-naive patients, but renin profiling is more useful for choosing add-on drugs.50
Empiric therapy with mineralocorticoid-receptor antagonist
Pharmacologic blockade of the mineralocorticoid receptor targets the excess aldosterone commonly present in patients with resistant hypertension.1,51 Spironolactone is most frequently used. Eplerenone, which has fewer sex-hormone–dependent adverse effects (e.g., painful gynecomastia and erectile dysfunction), is a costlier and less potent alternative.24 Amiloride, an epithelial sodium-channel blocker (not a mineralocorticoid-receptor antagonist), is another option and is used most commonly in primary aldosteronism.52
Dramatic reductions in office-based blood pressure readings (usually > 20 mm Hg) with add-on spironolactone treatment were reported in an observational study.53 In an RCT involving 117 patients with resistant hypertension, the addition of spironolactone reduced daytime ambulatory systolic blood pressure by 9.3 mm Hg, compared with a reduction of 3.9 mm Hg with placebo (p < 0.02).54 Corresponding reductions in office-based readings of systolic blood pressure were 14.6 mm Hg and 8.1 mm Hg (p = 0.01).54
Empiric therapy with sequential nephron blockade versus sequential renin–angiotensin system blockade
In a prospective, randomized, open-label trial involving 167 patients already taking irbesartan (300 mg/d), hydrochlorothiazide (12.5 mg/d) and amlodipine (5 mg/d), participants were randomly assigned to receive either sequential nephron blockade consisting of add-on therapy with low-dose diuretics (25 mg sprironolactone ± 20–40 mg furosemide ± 5 mg amiloride) or sequential renin–angiotensin system blockade (5–10 mg ramipril ± 5–10 mg bisoprolol) for 12 weeks.55 Renin profiling was not used in this study. Doses were adjusted or additional drugs added every four weeks. At 12 weeks, the mean 24-hour ambulatory blood pressure was lower in the group given sequential diuretic treatment (129/79 mm Hg v. 139/83 mm Hg; mean difference −10/−4 mm Hg; p < 0.01). These results show that empiric diuretic therapy is more likely than sequential renin–angiotensin system blockade to achieve blood pressure control in patients already receiving a standard base regimen.
Application in clinical practice
Box 3 illustrates the treatment of a patient with resistant hypertension.
Applying the results of this review in clinical practice
The following real case illustrates one approach to managing resistant hypertension. Alternative approaches are possible, and patient response may vary.
A 60-year-old woman with a history of diabetes, obesity, sleep apnea (treated with continuous positive airway pressure) and dyslipidemia is referred because of uncontrolled blood pressure. Her antihypertensive drugs are 20 mg lisinopril twice daily, 40 mg furosemide daily and 360 mg diltiazem (long-acting) daily. Other medications include insulin, atorvastatin, acetylsalicylic acid and metformin. The patient is a nonsmoker and does not consume alcohol. She follows a diet with no added salt and walks for 30 minutes three times a week. Her body mass index (BMI) is 34.1, and she weighs 78.7 kg. Her most recent hemoglobin A1c concentration was 8.6%, and she is followed by a diabetologist. Her serum creatinine level is 81 μmol/L, serum potassium level 5.0 mmol/L and low-density lipoprotein cholesterol 1.65 mmol/L. Her urinary albumin-to-creatinine ratio is normal, as were two previous 24-hour urine cortisol levels, the thyroid stimulating hormone level and the calcium level.
The automated blood pressure readings taken in the office are 165/78 mm Hg on average, and her systolic readings at home are frequently higher than 160 mm Hg. Twenty-four hour ambulatory blood pressure monitoring is offered to assess out-of-office readings more accurately, but the patient prefers to use her home blood pressure monitor. She is counselled on health behaviour modifications, including recommendations to lose weight and to increase her level of aerobic exercise to 60 minutes of walking most days of the week. Measurement of aldosterone and renin levels is requested.
Two weeks later, her average blood pressure is 158/67 mm Hg in the office and 156/78 mm Hg (over 12 measurements) at home. The need for health behaviour modifications is reinforced. Treatment with lisinopril and furosemide is replaced by a long-acting, once-daily combination preparation (8 mg perindopril and 2.5 mg indapamide) to reduce the pill burden and to simplify the dosing schedule.
One month later, her average blood pressure is 151/60 mm Hg in the office. Amlodipine is prescribed (5 mg once daily for two weeks and 10 mg once daily at bedtime thereafter), and diltiazem is stopped. (Switching from a dihydropyridine to a nondihydropyridine calcium-channel blocker eliminates the possiblility of excessive negative chronotropic action when a β-blocker is prescribed next).
After another month, her blood pressure levels are unchanged. Her aldosterone level is low normal (102 pmol/L), her renin level is over 10-fold above the laboratory normal limit, and her pulse rate is 85 beats/min. Because of the high renin level, the advantages, disadvantages and risks of screening for renal artery stenosis are discussed with the patient. She elects to undergo a renal scan with captopril (after a 48-hour hold of the ACE inhibitor). The scan is normal. Treatment with a long-acting anti-renin drug (5 mg bisoprolol once daily) is added.
Her average blood pressure in the office is 126/54 mm Hg one month later and 125/51 mm Hg two months later. Her average blood pressure levels at home are similar. She weighs 77.4 kg and has a BMI of 34.0. The importance of optimizing health behaviours is reinforced at her final visit before she is referred back to the care of her family physician.
What other treatment options are available?
Renal sympathetic denervation is a catheter-based percutaneous procedure (currently approved for use in Canada, Europe and elsewhere) that uses radiofrequency energy to ablate the afferent and efferent renal nerves (located in or adjacent to the arterial adventitial layer). The denervation procedure thereby targets elevated activity of the sympathetic nervous system and is based conceptually on surgical sympathectomy, a procedure used to treat resistant hypertension in the pre-antihypertensive era (~1920 to the 1950s).56 Reductions in central sympathetic outflow, renal vasoconstriction, renin–angiotensin activation, and sodium and water retention are the putative mechanisms involved in lowering blood pressure.
In a 6-month RCT involving 106 patients with resistant hypertension (systolic blood pressure ≥ 160 mm Hg or ≥ 150 mm Hg in patients with diabetes), blood pressure was reduced by 33/12 mm Hg (p < 0.0001) more in patients who underwent renal denervation than in the control group.57 Some nonrandomized studies have reported that renal denervation was associated with improvements in surrogate outcomes, such as sympathetic activity, left ventricular hypertrophy, glycemic control and diastolic dysfunction.58–61
Some of the reported criticisms of these studies are the lack of outcome data for 24-hour ambulatory blood pressure measurement and the lack of a sham procedure (catheterization without renal denervation) in the control group.58,62 In these prelininary studies, reductions in out-of-office blood pressure were much lower than the reductions in manual office-based readings, raising the possibility that a large portion of the reduction in blood pressure was unrelated to the procedure.58 A recently published, 6-month, single-blind, randomized, sham-controlled trial involving 535 patients reported no difference in the reduction of blood pressure between the renal denervation group and the control group (change in 24-hour ambulatory systolic pressure of −6.8 mm Hg v. −4.8 mm Hg, p= 0.98). 63 Office systolic blood pressure was also not different between the groups (difference 2.4 mm Hg, p= 0.26).
These findings cast uncertainty over the future use of renal denervation, and some manufacturers of denervation catheters have stopped any further studies pending internal reviews. The findings also emphasize the importance of performing careful, controlled assessment, including optimal outcome assessment, before the widespread adoption of new technologies. Although the results could be explained theoretically by incomplete or ineffective renal denervation, improved adherence by patients to background drug therapy, regression to the mean, a “placebo” effect of the sham procedure, and additional factors are as or more likely.64 Overall, renal denervation is of uncertain efficacy at this time, and its use cannot be recommended outside of ongoing and future research trials.
Unanswered questions
The nonpharmacologic and pharmacologic management of resistant hypertension is largely based on consensus recommendations by experts. Algorithm-based approaches, such as renin profiling to guide drug selection, require further validation. Comparative effectiveness RCTs are needed to identify the most efficacious treatment regimens.
New drugs designed to counteract vasoconstriction, fibrosis and inflammation, to inhibit aldosterone synthesis or to reduce arterial stiffness are under investigation; however, early efficacy results for many of these drugs have been disappointing, and unacceptable adverse events have occurred.65 Device-based treatments are under active and intense investigation: renal sympathetic denervation and carotid baroreflex stimulation are the furthest along the development path. Increased uptake will likely occur if the long-term safety and efficacy of these treatments are established.
KEY POINTSResistant hypertension is estimated to affect 10% of adults with hypertension.
Pseudoresistance and secondary hypertension must be excluded.
The optimal base regimen for most patients comprises a diuretic, an angiotensin-converting-enzyme inhibitor or angiotensin-receptor blocker, and a dihydropyridine calcium-channel blocker.
Long-acting combination products should be used to maximize adherence to treatment.
Mineralocorticoid receptor antagonists, and α- and β-blockers are the most commonly used add-on drugs.
Footnotes
Competing interests: Raj Padwal has received speaker or consultant fees from Forest Labs, Medtronic, Abbott Labs and Servier. He has conducted clinical trials with Novo Nordisk and CVRx. No competing interests were declared by Simon Rabkin and Nadia Khan.
This article has been peer reviewed.
Contributors: Raj Padwal and Nadia Khan wrote the initial draft of the manuscript. All of the authors critically revised the manuscript, approved the final version submitted for publication and agreed to act as guarantors of the work.