Current Hypertension Reports

, 13:347

Resistant Hypertension: Concepts and Approach to Management


  • Gary E. Sander
    • Heart and Vascular InstituteTulane University School of Medicine
    • Heart and Vascular InstituteTulane University School of Medicine

DOI: 10.1007/s11906-011-0226-7

Cite this article as:
Sander, G.E. & Giles, T.D. Curr Hypertens Rep (2011) 13: 347. doi:10.1007/s11906-011-0226-7


Resistant hypertension (RH), defined simply, is blood pressure (BP) requiring the use of four or more antihypertensive agents, whether controlled or uncontrolled. RH is an increasingly common problem in elderly patients and may affect as many as 20% of the hypertensive population. Unfortunately, at least 30% of patients evaluated for RH are actually adequately controlled when more carefully assessed by home BP monitoring or ambulatory BP monitoring, thus representing a white coat effect. It is also essential to exclude pseudoresistance resulting from improper BP recording techniques or failure of the patient to adhere to the prescribed treatment regimen. Concurrent use of drugs that may interfere with prescribed antihypertensive agents, including many over the counter herbal preparations, must also be excluded. The underlying mechanisms principally driving true RH include pathophysiologic abnormalities of aldosterone signaling, sodium and water retention, excessive sympathetic nervous system activity, and obstructive sleep apnea. Appropriate treatment regimens will usually include an inhibitor of the renin-angiotensin-aldosterone system, a calcium channel blocker, and a diuretic. An aldosterone receptor blocker can be instituted at any step, and is very effective as a fourth drug. Beta-blockers can also be integrated into these treatment plans and may be especially helpful when excessive sympathetic nervous system activity is suspected. Novel device therapies that interrupt sympathetic nerve stimulation at the carotid sinus and kidney are under investigation, and may add entirely new directions in the management of RH. What is most important is that treatment regimens should be targeted to specific patient profiles.


Resistant hypertensionUncontrolled hypertensionBlood pressureMechanismsManagement: TreatmentMonitoringWhite coat effectPseudoresistanceAntihypertensive agentsObstructive sleep apneaSympathetic nervous system activitySpironolactone


Blood pressure is the principal clinical biomarker of hypertension and is used to chart the progress of the control of the disease. Put simply, “resistant hypertension” (RH) or “uncontrolled hypertension” is difficult-to-control BP in a person with hypertension. This entity is not a trivial problem: it describes perhaps 20% of the adult hypertensive population. Efficient and correct assessment and treatment will often require consultation with a hypertension specialist. This review is intended to assist in identifying correctable causes of resistant hypertension and recommending treatment strategies.


The Seventh Report of the Joint National Committee 7 (JNC7) defined resistant hypertension as lack of control of blood pressure (BP) when the treatment regimen has included three or more drugs of different classes, one of which was a diuretic, at appropriate doses [1]. This definition was further refined in an American Heart Association guideline paper on the diagnosis and treatment of resistant hypertension [2]. This publication defined RH as BP that remains above goal despite the use of three antihypertensive drugs of different classes. Ideally, one of the three drugs should be a diuretic, and all agents should have been prescribed at optimal doses. RH also includes patients whose BP is controlled but who require more than three medications; these patients should be considered resistant to treatment. A simplified definition is that RH is BP requiring the use of at least four antihypertensive agents, whether controlled or uncontrolled. An additional category of “refractory hypertension” has been suggested for patients who do not reach target BP despite using at least four medications. As will be discussed later, this category may be helpful because its underlying mechanisms (and therefore its treatment) appear to be different from those of RH. For the remainder of this discussion, RH will indicate only those patients meeting the criteria for resistant, rather than refractory, hypertension.


RH is a common and increasingly prevalent problem, identifying patients at high cardiovascular risk but with a high incidence of curable causes of hypertension (including secondary hypertension), who may benefit from special therapeutic and diagnostic considerations. Furthermore, it is a functional definition that can be applied for both clinical and research applications. In a recent analysis of National Health and Nutrition Examination Survey (NHANES) participants being treated for hypertension, only 53% were controlled to BP less than 140/90 mm Hg [3]. In patients with chronic kidney disease (CKD), only 37% were controlled to less than 130/80 mm Hg [4], and only 25% of diabetic participants were controlled to less than 130/85 mm Hg [3]. Results from the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT) demonstrated that 27% of participants were using three or more medications after approximately 5 years of follow-up [5]. Data from a large, community-based practice network (n = 264,967) indicated that 12.7% of hypertensive patients were uncontrolled on three or more medications, and 3.5% on at least four medications [6]. Changes in the aggressiveness of BP treatment from 1993 to 2004 have been reported [7]. In 1993, 52% of the study population received monotherapy and 48% received multidrug regimens; by 2004, only 40% received monotherapy and 60% received multiple drugs.

Patient Characteristics

Clinical characteristics identified in RH patients include older age, high baseline BP, obesity, sleep disorders, excessive dietary salt ingestion, chronic kidney disease, diabetes, left ventricular hypertrophy, black race, female sex, and residence in the southeastern United States. Odds ratios for achieving BP control with some of these factors are older age, 0.92; obesity, 0.87; female sex, 0.83; high baseline BP, 0.82; chronic kidney disease, 0.79; left ventricular hypertrophy, 0.69; and black race, 0.69 [5]. Lower serum potassium levels [8] have also been described. Patients with chronic kidney failure [9, 10] and renal artery stenosis [11] constitute special groups in whom control of BP may be truly difficult; these issues have been reviewed elsewhere and will not be discussed here.

Clinical Evaluation of Resistant Hypertension

True RH must first be distinguished from pseudoresistance due to patient nonadherence to prescribed drug regimens, improper BP measurement techniques, inadquate dosing of prescribed drugs, inappropriate drug combinations, and the “white coat effect.” The frequency of inadequate patient compliance with medications cannot be underestimated. When monitoring of compliance using electronic pillboxes was offered to patients, BP was normalized in one third of the patients, and insufficient compliance was unmasked in another 20% [12]. The technique of BP measurement is critically important [13]. BP should be measured after a patient has been seated quietly for 5 min, with his or her arm supported at heart level and with the use of a properly calibrated and sized cuff; inappropriate cuff size may erroneously increase the systolic BP measurement by as much as 5 to 15 mm Hg. The patient should not have smoked a cigarette within the previous 15 to 30 min, as smoking can elevate systolic BP by 5 to 20 mm Hg. Avoidance of coffee is also recommended, although the increase in systolic BP after one cup of caffeinated coffee is usually only 1 to 2 mm Hg.

On the other hand, white coat hypertension, defined as persistently high office BP with normal or controlled self-measured or ambulatory BP, is much more prevalent in patients with RH, accounting for as many as one third of patients with RH versus 12% in the non-RH hypertensive population [14]; at least 30% of patients labeled as resistant may fall into this category. Data collected on 68,045 patients from the 2009 Spanish Ambulatory BP Monitoring Registry identified 8,295 people as having RH. After ambulatory BP monitoring (ABPM), 62.5% of these patients were classified as truly having RH; this group was younger, more frequently men, and had a longer duration of hypertension and a worse cardiovascular risk profile. They included larger proportions of smokers, diabetics, and patients with target organ damage (including left ventricular hypertrophy, impaired renal function, and microalbuminuria) and documented cardiovascular disease. The remaining 37.5% of patients were, in fact, adequately controlled [14]. A smaller RH group was evaluated by ABPM using two different cutoff levels: daytime mean ambulatory BP less than 135/85 mm Hg and 24-hour mean ambulatory BP less than 130/80 mm Hg. The percentage of patients with white coat effect was identified as 31% by daytime cutoff parameters and 25% by 24-hour parameters; 67% of patients were truly resistant according to both cutoff points [15]. Thus the white coat effect may occur at least at initial evaluation in roughly one third of patients with suspected RH and should be strongly considered, particularly in patients who lack evidence of target-organ damage or who have hypotensive episodes after counseling for the importance of compliance with medications. ABPM or at least home monitoring may be necessary to differentiate true RH from a white coat effect.

In elderly patients, “pseudohypertension,” defined as falsely high BP measurements resulting from vascular stiffening, should be considered in similar circumstances of hypotension or absent target-organ damage. Palpating the radial artery while simultaneously occluding the brachial artery with the cuff (the Osler maneuver) may be helpful in recognizing this process [13, 16].

It is imperative to consider whether prescribed drugs or over the counter herbal preparations may be causing elevations in BP or interfering with the actions of antihypertensive drugs (Table 1). Drug-induced hypertensive crises have been reported [1720].
Table 1

Agents that can increase blood pressure

Anabolic steroids

Antidepressants, particularly venlafaxine, tricyclics, MAO inhibitors

Anti-obesity drugs, particularly OTC combinations that include ephedra



Herbal preparations, particularly ephedra


Mineralocorticoids, including licorice


Oral contraceptives containing estrogen

Serotonergics, including antimigraine drugs such as sumatriptan


Stimulants, particularly amphetamines and derivatives


Sympathomimetic amines

Coxibs COX-2 inhibitors; MAO monoamine oxidase; NSAIDs nonsteroidal antiinflammatory drugs; OTC over the counter

After confirmation of true RH, patients should be evaluated for the presence and severity of target-organ damage, including left ventricular hypertrophy, retinopathy, chronic kidney disease, and proteinuria, and should be screened for secondary causes of hypertension, including primary aldosteronism, renal artery stenosis, and obstructive sleep apnea (OSA) (Table 2). Less common but possible causes include pheochromocytoma, Cushing’s disease, hyperparathyroidism, aortic coarctation, and intracranial tumor [1, 2].
Table 2

Diagnostic characteristics of secondary causes of hypertension


Diagnostic characteristics

Obstructive sleep apnea

• Snoring, witnessed apnea, excessive daytime sleepiness

Primary aldosteronism

• Hypokalemia

• ↑ PAC

• ↓ Plasma renin activity (PRA)

• PAC/PRA ratio ≥ 20 ng/dL per ng/mL per hour (≥ 555 pmol/L per ng/mL per hour) and

• Elevated PAC ≥ 15 ng/dL (≥ 416 pmol/L)

Chronic kidney disease

• Creatinine clearance < 30 mL/min

Renal artery stenosis

• Azotemia induced by ACE inhibitors or ARBs

• Abdominal bruit, particularly in diastole

• Continuous epigastric bruit

• Diffuse atherosclerotic vascular disease (carotid, coronary, peripheral, aorta)

• Flash pulmonary edema

• Unexplained worsening of coronary ischemia

• Unexplained worsening of heart failure

• Unilateral small kidney (any radiologic study)


• Episodic hypertension, palpitations, diaphoresis, headache

Cushing’s disease

• Moon facies, central obesity, abdominal striae, interscapular fat deposition

Aortic coarctation

• Differential in brachial or femoral pulses; systolic bruit

ACE angiotensin-converting enzyme; ARB angiotensin receptor blocker

Mechanisms of Resistant Hypertension

It is critical to understand the pathophysiological processes that lead to the development of RH in order to better design effective antihypertensive treatment algorithms. Hyperaldosteronism, salt sensitivity, and disordered sleep (especially OSA) have been implicated. Aldosterone excess, OSA, and RH commonly coexist, perhaps related by fluid retention [21].


There is some confusion in the biochemical evaluation of aldosterone excess. Authors have used differing normal ranges, particularly differing plasma aldosterone-to-renin ratios (ARR), so caution is warranted in comparing studies. A number of studies have demonstrated abnormal aldosterone concentrations in RH, and there is a relationship between the severity of hypertension and the extent of aldosterone concentrations. Hyperaldosteronism, defined as an ARR greater than 25 and confirmed by the fludrocortisone test, was detected in 63 of 609 hypertensive patients [22]. When these patients were stratified according to hypertension stage (JNC VI classification) [23], prevalence increased by stage: stage 1, 6/301 cases (1.99%); stage 2, 15/187 cases (8.02%); and stage 3, 16/121 cases (13.2%). Serum potassium levels were normal in 36 of 37 patients; one patient had minor hypokalemia. CT scans showed bilateral adrenal enlargement in seven patients and an adrenal nodule in 2. The low frequency of CT scan abnormalities and hypokalemia suggests that the diagnosis for most primary hyperaldosteronism patients corresponds to attenuated forms of the disease [22]. Another study compared 279 consecutive RH patients with 53 controls (normotensive individuals or patients with controlled BP using no more than two antihypertensive medications) [8]. While on a normal diet, the patients with RH had higher values for plasma aldosterone (PAldo) (P < 0.001), ARR (P < 0.001), 24-hour urinary aldosterone (UAldo) (P = 0.02), brain-type natriuretic peptide (BNP) (P = 0.007), and atrial natriuretic peptide (ANP) (P = 0.001), and lower values for plasma renin activity (PRA) (P = 0.02) and serum potassium (P < 0.001). Although only 22% had elevated PAldo concentrations (>20 ng/mL), 60% had PRA levels less than 1.0 ng/mL per hour and 35% had ARR greater than 20. Men with RH had significantly higher values than women for PAldo (P = 0.003), ARR (P = 0.02), 24-hour UAldo (P < 0.001), and urinary cortisol (P < 0.001). In univariate linear regression analysis, values correlated with 24-hour UAldo levels included body mass index (P = 0.01), serum potassium (P < 0.001), urinary cortisol (P < 0.001), urinary sodium (P = 0.02), and urinary potassium (P < 0.001). Levels of serum potassium (P = 0.001), urinary potassium (P < 0.001), and urinary sodium (P = 0.03) were predictors of 24-hour UAldo levels in multivariate modeling. PAldo levels are higher in patients with RH, and there is evidence of intravascular volume expansion (higher BNP/ANP levels); differences are most pronounced in men. A significant correlation between 24-hour UAldo levels and cortisol excretion suggests that a common stimulus, such as corticotropin, may underlie the aldosterone excess in RH.

Another prospective study from the same university clinic measured 24-hour UAldo during high dietary salt ingestion, baseline PRA, and PAldo in all 88 consecutive RH patients. Primary hyperaldosteronism was confirmed if PRA was less than 1.0 ng/mL per hour and UAldo was greater than 12 mcg per 24 h during high urinary sodium excretion (>200 mEq/24 h). The accuracy of these results was confirmed by suppression testing. Primary hyperaldosteronism was confirmed in 18 patients (20%), with similar occurrence in both black and white patients. Of the 14 patients with confirmed hyperaldosteronism who were treated with spironolactone, all manifested a significant reduction in BP, supporting the importance of aldosterone in treatment resistance. In this population, an elevated ARR (>20) had a sensitivity of 89% and a specificity of 71%, with a corresponding positive predictive value of 44% and a negative predictive value of 96% [24]. Further studies have supported the suggestion that the ARR is a valid screening assay for primary aldosteronism in patients with poorly controlled BP, and discontinuation of antihypertensive medications is not needed for this test [25]. In four international studies, the incidence of primary hyperaldosteronism has been reported to be between 17% and 22% [2427].

The use of aldosterone antagonists may result in an increase in serum potassium levels. To assess the safety and efficacy of aldosterone antagonists in the presence of stage 2 or stage 3 CKD (mean estimated glomerular filtration rate [eGFR] of 56.5 ± 16.2 mL/min/1.73 m2), 46 RH patients with CKD received three mechanistically complementary antihypertensive agents, including a diuretic and a renin-angiotensin-aldosterone system (RAAS) blocker, and spironolactone was added to their treatment regimens. The mean increase in serum potassium was 0.4 mEq/L above baseline (P = 0.001), with 17.3% experiencing an increase in serum potassium of more than 5.5 mEq/L. Predictors of hyperkalemia included a baseline eGFR of 45 mL/min/1.73 m2 or higher, with serum potassium greater than 4.5 mEq/L on appropriately dosed diuretics [28].

Obstructive Sleep Apnea

About 50% to 60% of patients with OSA have hypertension. The prevalence of OSA, defined as an apnea-hypopnea index (AHI) of at least ten obstructive events per hour of sleep, was 83% in 24 men and 17 women with RH who were studied. Patients were generally late middle-aged (57.2 ± 1.6 years), predominantly white (85%), obese (body mass index, 34.0 ± 0.9 kg/m2), and taking a mean of 3.6 ± 0.1 different antihypertensive medications daily. OSA was more prevalent in men than in women (96% vs 65%, P = 0.014) and was more severe in men (mean AHI, 32.2 ± 4.5 vs 14.0 ± 3.1 events per hour; P = 0.004). There was no gender difference in body mass index or age. Women with OSA were significantly older and had a higher systolic BP, lower diastolic BP, wider pulse pressure, and slower heart rate than women without OSA [29]. There does appear to be a relationship between hyperaldosteronism and OSA. The Kaiser Permanente Southern California database has been searched to identify hypertensive individuals with hyperaldosteronism (defined as an ARR > 30 and PAldo > 20 ng/mL or an ARR > 50). Of 3,428 hypertensive patients, 575 (17%) had hyperaldosteronism, and OSA was present in 18% with hyperaldosteronism versus 9% without hyperaldosteronism (P < 0.001); the odds ratio for OSA in patients with hyperaldosteronism was 1.8 (95% CI, 1.3–2.6) after controlling for other OSA risk factors. All ethnic groups had similar risk [30].

From yet another series of 109 consecutive RH patients evaluated for hyperaldosteronism with measurement of PRA, PAldo, 24-hour UAldo excretion, and polysomnography, hyperaldosteronism was found in 28% and OSA in 77% [31]. In the presence of hyperaldosteronism, OSA prevalence was 84%, compared with 74% in hypertensive patients with normal aldosterone levels. PAldo and UAldo were both significantly correlated with the AHI in the high-aldosterone group, but there were no significant differences in body mass index or neck circumference between aldosterone groups. These observations confirmed both a markedly high prevalence of OSA in patients with RH and a greater severity of OSA in those patients with demonstrable hyperaldosteronism, further related to the degree of aldosterone excess. These findings support the hypothesis that aldosterone excess contributes to greater severity of OSA [31].

Salt Sensitivity

The role of salt sensitivity in the development of RH was evaluated in a study that examined the effects of dietary salt restriction on office BP measurements and 24-hour ABPM in individuals with RH [32]. Twelve RH patients (using 3.4 ± 0.5 drugs, with mean office BP of 145.8 ± 10.8/83.9 ± 11.2 mm Hg) were entered into a randomized crossover evaluation of a low-sodium diet (50 mmol/24 h × 7 days) and a high-sodium diet (250 mmol/24 h × 7 days) separated by a 2-week washout period. BNP, PRA, 24-hour UAldo, aldosterone, sodium, potassium, 24-hour ABPM, aortic pulse wave velocity, and augmentation index were compared between dietary treatment periods. Compared with the high-salt diet, the low-salt diet decreased office systolic BP by 22.7 mm Hg and diastolic BP by 9.1 mm Hg. PRA increased during low salt intake, whereas BNP and creatinine clearance decreased, indicative of intravascular volume reduction, demonstrating that ingestion of excessive dietary sodium contributes importantly to resistance to antihypertensive treatment. Again, an interaction among risk factors is demonstrated: high PAldo and high dietary sodium intake combine to worsen proteinuria in patients with RH, and a positive correlation between urinary protein and urinary sodium was observed in patients with high UAldo but not in those with normal UAldo [21].


Although adjusted logistic regression has found no significant association between the studied polymorphisms and controlled or resistant hypertension, multifactor dimensionality reduction analyses have shown that carriers of the AGT 235 T (angiotensinogen) allele are at increased risk for RH, especially if they are older than 50 years. The AGT 235 T allele constitutes an independent risk factor for RH [33].

Treatment Considerations


The role of spironolactone in the management of RH has been well demonstrated. Eplerenone, a more specific blocker of the aldosterone receptor, has fewer adverse effects than spironolactone and may be used in patients with spironolactone intolerance [34]. However, no data have actually demonstrated the efficacy of eplerenone in RH. In 76 subjects, 34 of whom had biochemical primary hyperaldosteronism, low-dose spironolactone (12.5 to 25 mg/day) was associated with an additional mean decrease in BP of 21 ± 21/10 ±14 mm Hg at 6 weeks and 25 ± 20/12 ± 12 mm Hg at 6-month follow-up. The BP reduction was similar in patients with and without primary hyperaldosteronism and was additive to the use of angiotensin-converting enzyme (ACE) inhibitors, angiotensin receptor blockers (ARBs), and diuretics. In fact, systolic BPs trended to greater reductions in the so-called non–primary hyperaldosteronism group [35]. The Anglo-Scandinavian Cardiac Outcomes Trial–BP Lowering Arm (ASCOT-BPLA) used spironolactone mainly as a fourth-line add-on therapy at the discretion of investigators, when study medication (amlodipine/perindopril versus atenolol/bendroflumethiazide with doxazosin as the third drug) had failed; spironolactone doses between 25 and 50 mg daily reduced systolic BP by 21.9 mm Hg and diastolic BP by 9.5 mm Hg [36]. The efficacy of BP reduction with spironolactone has been confirmed by ABPM [37]; in this study, the factors associated with better response were higher waist circumference, lower aortic pulse wave velocity, and lower serum potassium. No association with PAldo or ARR ratio was found.

Treatment with spironolactone has been compared with dual RAAS blockade using an ACE inhibitor and an ARB. Following the first stage of dual RAAS blockade, systolic BP dropped significantly both in the office (−12.9 ± 19.2 mm Hg) and by ABPM (−7.1 ± 13.4 mm Hg). Diastolic BP was unchanged in the office but was significantly reduced when measured by ABPM (−3.4 ± 6.2 mm Hg). Spironolactone reduced office BP by 32.2 ± 20.6/10.9 ± 11.6 mm Hg and ABPM by 20.8 ± 14.6/8.8 ± 7.3 mm Hg (P < 0.001). Office BP control was achieved by 25.6% of patients on dual blockade and 53.8% using spironolactone. By ABPM, 20.5% were controlled on dual blockade and up to 56.4% with spironolactone. Thus, spironolactone at daily doses of 25 to 50 mg had a greater antihypertensive effect than dual blockade of the RAAS in patients with RH [38].


In the absence of significant renal insufficiency, thiazides are preferred over loop diuretics, the chronic use of which may result in distal tubular cell hypertrophy [39]. If a loop diuretic is required, torsemide may be preferable to furosemide for single daily dosing [40]. There have been reports that, within recommended doses, chlorthalidone (with a plasma half-life of 45–60 h) is more effective in lowering systolic BP than hydrochlorothiazide, as evidenced by 24-hour ABPM [41, 42]; there is considerable debate about whether hydrochlorothiazide should be used at all. However, chlorthalidone has been demonstrated to cause sympathetic nervous system activation (SNA) [43], which may contribute to poorer outcomes in patients with cardiovascular disease. Compared with spironolactone, chlorthalidone caused a similar reduction in ABPM but increased SNA by 23% (P < 0.01) within 1 month of treatment, an effect persisting after 3 months; spironolactone had no SNA effect in the same patients. Baroreflex control of SNA was unaffected by either drug. Chlorthalidone increased indices of insulin resistance, an effect significantly correlated with increases in SNA, and spironolactone again had no effect in the same patients. Adding spironolactone to chlorthalidone reversed this SNA activation, suggesting that combining spironolactone with chlorthalidone may reduce the adverse effects of chlorthalidone [44].

Amiloride and hydrochlorothiazide combination therapy has also been demonstrated to be effective. Of 60 low-renin RH patients (PRA < 0.5 nmol/L per hour), a fixed combination of amiloride and hydrochlorothiazide reduced BP by 31/15 mm Hg compared with placebo (P ≤ 0.0001). Serum aldosterone and plasma renin activity increased substantially during active treatment [26].

Combination Therapies

There are many options and opinions as to the crafting of treatment regimens for RH; the major consideration is that they should be targeted to individual patients. General algorithms have been proposed. One such interesting proposal advocates a mechanism-based, multiple-drug regimen targeting treatment at the three hypertensive mechanisms: sodium/volume, RAAS, and SNA [45]. Such a regimen starts with baseline treatment with a two-drug combination directed at sodium/volume and the RAAS. The recommendation is to proceed with one or both of just two treatment options: (1) strengthening the diuretic regimen, possibly with the addition of spironolactone, and/or (2) adding agents directed at the SNA, usually a β-blocker or a combination of an α-blocker and a β-blocker. However, the results of the ACCOMPLISH trial (Avoiding Cardiovascular Events through Combination Therapy in Patients Living with Systolic Hypertension) suggest that amlodipine may offer benefit over thiazides as a first addition to RAAS manipulation [46]. In this trial of benazepril/amlodipine versus benazepril/hydrochlorothiazide, the amlodipine arm was shown to be more effective than the hydrochlorothiazide arm in reducing cardiovascular event rates. Further analysis indicated that 24-hour BP control (mean systolic BP <130 mm Hg on ABPM) was similar for both treatment arms, supporting the interpretation that the difference in cardiovascular outcomes favoring an RAAS blocker combined with amlodipine rather than hydrochlorothiazide was not caused by differences in BP, but rather by intrinsic properties (metabolic or hemodynamic) of the combination therapies [47]. The possibility of a negative effect of thiazides in the reduction of cardiovascular events has been suggested [48]. Furthermore, it seems prudent to consider spironolactone as an aldosterone receptor antagonist rather than as an additive diuretic.

Thus it seems reasonable that most patients should initially be treated with three-drug combination therapies including an RAAS inhibitor, a calcium channel blocker, and then a diuretic. There is probably little difference to be anticipated among the agents within the RAAS inhibitor class. This may not be true of the calcium channel blocker class: first-generation dihydropyridine (DHP) calcium channel blockers are known to increase albuminuria. In RH patients treated to similar levels of BP reduction with an ACE inhibitor or ARB, a diuretic, and at least a third drug in the highest tolerated dose, albumin excretion was 14.3 ± 16.9 mg/d with DHP and 8.5 ± 6.6 mg/d in patients treated without DHP (P = 0.036). Thus, even in patients treated with an ACE inhibitor or ARB, DHP calcium channel blockers are associated with increased albuminuria, which is associated with increased cardiovascular risk [49], suggesting that non-DHP or later-generation calcium channel blockers may be preferable. In CKD patients with proteinuria, RAAS inhibitors should be used as first-line antihypertensive therapy because these agents seem to have a BP-independent antiproteinuric effect. If target BP levels are not achieved, a diuretic should be added to this regimen. A combination of an ACE inhibitor with an ARB or other classes of medications shown to decrease protein excretion, such as a non-DHP calcium channel blocker or spironolactone, should be considered to decrease proteinuria further [50]. However, the ACE inhibitor/ARB combination should be used cautiously, as it may produce excessive BP reduction and worsen renal function [51].

Chronotropic Considerations

Therapeutic strategies in RH have included adding or changing drugs in a search for a better synergic combination. Most patients receive all medications in a single morning dose, ignoring the potential of the circadian pattern. To investigate the importance of circadian patterns, 250 RH patients receiving three antihypertensive drugs in a single morning dose were randomly assigned to either changing one drug and continuing to take all three in the morning or making the same change but administering the new drug at bedtime. There was no effect on ambulatory BP when all of the drugs were taken on awakening, but a statistically significant reduction occurred in those taking the new drug at bedtime (−9.4/6.0 mm Hg; P < 0.001). The reduction was larger in the nocturnal than in the diurnal mean BP. Only 16% of the patients in the group administered the drug at bedtime were dippers at baseline, compared with 57% after therapy (P < 0.001), suggesting that time of treatment may be more important than the drug combination in controlling the circadian BP pattern [52].

Among patients with diabetes, treatment with one or more hypertension medications at bedtime, compared with all medications upon waking, similarly improved ambulatory BP control and significantly reduced cardiovascular risk by 12% for each 5 mm Hg decrease in asleep systolic BP during follow-up (P < 0.001) [53].

Also related to circadian issues is the specific situation of OSA. Apneas stimulate ANP release and SNA, effects that persist throughout the day. The combination of increased SNA and nocturnal diuresis help to explain observations that β-1 antagonists lower BP more than thiazide diuretics in patients with OSA; ACE inhibitors and ARBs have been equally effective, with spironolactone effective as a fourth drug. It may be reasonable to prescribe antihypertensive drugs at night to treat nocturnal hypertension [54].

New Drugs

Drugs in development may add to the antihypertensive regimen in RH. Darusentan, a new vasodilatory, selective endothelin type A antagonist, reduced BP in the clinic in patients who were receiving at least three antihypertensive drugs, including a diuretic, at full or maximum tolerated doses. The reductions were 17/10 ± 15/9 mm Hg at a dose of 50 mg daily, 18/10 ± 16/9 mm Hg with 100 mg daily, and 18/11 ± 18/10 mm Hg with 300 mg daily, compared with placebo (9/5 ± 14/8 mm Hg; P < 0.0001 for all). The main adverse effects were related to fluid accumulation. As with other vasodilatory drugs, edema was noted, occurring in 27% of the 247 patients given darusentan, compared with 14% of the 132 who were given placebo [55].

Device Therapy

Baroreflex activation therapy involves implantation of the Rheos Baroreflex Hypertension Therapy System into the neck around the carotid sinus, with subsequent activation of the carotid baroreflex through electrical stimulation of the carotid sinus wall. Studies in both normotensive and hypertensive canine models have demonstrated sustained and clinically relevant reductions in BP and SNA with prolonged baroreflex activation [56, 57]. An initial clinical trial has reported encouraging results [58]. In this trial, 45 patients with systolic BP of at least 160 mm Hg or diastolic BP at least 90 mm Hg despite at least three antihypertensive drugs were enrolled in a prospective, nonrandomized feasibility study. Baseline mean BP was 179/105 mm Hg, and heart rate was 80 beats per minute despite a median of five antihypertensive drugs. After 3 months of device therapy, mean BP was reduced by 21/12 mm Hg, a result sustained in 17 patients who completed 2 years of follow-up, with a mean reduction of 33/22 mm Hg.

A proof-of-principle trial of therapeutic renal sympathetic denervation using percutaneous radiofrequency catheter-based treatment was conducted in 50 RH patients. Baseline mean office BP was 177/101 ± 20/15 mm Hg on an average of 4.7 antihypertensive medications. After renal sympathetic denervation, BP was reduced by 14/10 mm Hg at 1 month, 21/10 mm Hg at 3 months, 22/11 mm Hg at 6 months, 24/11 mm Hg at 9 months, and 27/17 mm Hg at 12 months [59].

Refractory Hypertension

It has been suggested that there is a unique phenotype of patients with refractory hypertension, defined as BP that remains uncontrolled in spite of maximal multidrug therapy (Calhoun DA, 2010, personal communication). The prevalence of refractory hypertension may be as high as 15% of patients referred to specialty clinics for RH. Unlike RH, biochemical analysis and lack of response to spironolactone argues against primary aldosteronism as the underlying mechanism of treatment resistance. Furthermore, normal BNP levels that are unchanged with diuretic therapy argue against volume dependence. Elevated heart rate, especially after β-blocker therapy, suggests possible heightened sympathetic tone.


RH is an increasingly common problem that requires careful attention in order to design appropriate treatment protocols and thus reduce the significant cardiovascular risks associated with hypertensive cardiovascular disease. In evaluating patients with RH, it is first critical to exclude “pseudoresistance” due to such factors as incorrect BP measurement, patient noncompliance with treatment regimens, interfering drugs or herbal preparations, and white coat hypertension. True RH is driven principally by underlying mechanisms that include abnormities of aldosterone signaling, sodium and water retention, excessive SNA, and OSA. Appropriate treatment regimens can then be targeted to specific patient profiles. These regimens will usually include an RAAS inhibitor, a calcium channel blocker, and a diuretic. An aldosterone receptor blocker can be instituted at any step.


Conflicts of Interest: G. Sander: Fees from Forest Labs related to their speaker program.

Copyright information

© Springer Science+Business Media, LLC 2011