David A. Calhoun, MD, 430 BMR2, 1530 3rd Avenue South, Birmingham, AL 35294–2180 E-mail: firstname.lastname@example.org
Treatment of resistant hypertension requires confirmation of true resistance, diagnosis and treatment of secondary causes of hypertension, adoption of appropriate lifestyle modifications, and effective use of multidrug antihypertensive regimens. Excessive volume retention often underlies resistant hypertension, so diuretics are generally necessary to achieve blood pressure (BP) goals. Although treatment regimens consisting of 3 or more agents have not been systematically evaluated, the author has found a triple regimen consisting of a thiazide diuretic, a calcium channel blocker, and an angiotensin-converting enzyme (ACE) inhibitor or angiotensin receptor blocker (ARB) to be generally effective and well tolerated. Although hydrochlorothiazide is more widely used, chlorthalidone provides better BP reduction and should be preferentially used in patients with resistant hypertension, particularly if the patient remains uncontrolled on hydrochlorothiazide. Recent studies have demonstrated that low doses of aldosterone antagonists, when added to multi-drug regimens that include a thiazide diuretic and an angiotensin-converting enzyme inhibitor or angiotensin receptor blocker, provide significant additional BP reduction, seemingly exceeding what would be expected with addition of alternative classes of agents. The degree of BP reduction induced by aldosterone blockade has been similar in patients with and without evidence of aldosterone excess. Aldosterone antagonists are generally safe and well tolerated. The most common adverse effect of low-dose spironolactone has been breast tenderness, occurring in about 10% of men. Hyperkalemia is uncommon, but can occur, necessitating biochemical monitoring. Risk of hyperkalemia is increased in patients with chronic kidney disease or diabetes, elderly patients, and patients already receiving an angiotensin-converting enzyme inhibitor or angiotensin receptor blocker.
Treatment of resistant hypertension requires: (1) exclusion of pseudoresistance, (2) identification and correction of contributing factors and/or secondary causes of hypertension, (3) adoption of appropriate lifestyle changes, and (4) use of effective multidrug antihypertensive regimens.1 The most common causes of pseudoresistant hypertension include nonadherence to the prescribed pharmacologic regimen, inaccurate blood pressure (BP) measurement technique resulting in falsely high BP readings, and white coat resistant hypertension. Lack of adherence to prescribed antihypertensive medications is probably the most common cause of failure to control BP.1–3 Such a situation, however, is not true resistant hypertension, as a regimen cannot be considered ineffective if it has not been followed correctly. Ultimately, physicians must rely on patient self-report to accurately assess the degree of adherence. To facilitate accurate self-report, physicians must maintain good rapport with patients and attempt to identify in a nonjudgmental fashion barriers preventing full adherence—including adverse effects, out-of-pocket expenses, or difficulties with the dosing regimen. Reducing the number of prescribed pills by using fixed-dose combinations, simplifying the dosing regimen by using once-daily preparations, and reducing costs incurred by the patient have all been shown to improve patient adherence.
Other causes of apparent treatment resistance include inaccurate BP measurements due to poor technique and significant white coat effects. Poor BP technique often includes use of too small a BP cuff and/or not allowing the patient to sit quietly for 3–5 minutes before checking the BP, both of which may result in falsely high BP readings. A significant white coat effect will result in high office BP readings but lower BP levels out of the office. Ambulatory BP monitoring or home and office BP assessments can be used to identify white coat resistant hypertension, ie, uncontrolled BP limited to office measurements.
Lifestyle factors contributing to resistant hypertension are common and include obesity, physical inactivity, and high dietary salt intake. Accordingly, recommendations of weight reduction, exercise, and reductions in dietary salt ingestion are appropriate for patients with resistant hypertension. Although not specifically evaluated in patients with resistant hypertension, a high-fiber (fruits and vegetables), low-fat, and low-sodium diet (ie, Dietary Approaches to Stop Hypertension [DASH] diet) has been demonstrated to lower BP in hypertensive populations.4
The prevalence of secondary causes of hypertension is increased in patients with resistant hypertension; in particular, primary aldosteronism (PA), obstructive sleep apnea, chronic kidney disease (CKD), and renal artery stenosis.5–7 All patients confirmed to have resistant hypertension should be screened for these secondary causes, with referral to an appropriate specialist as needed.
PHARMACOLOGIC TREATMENT RECOMMENDATIONS
Pharmacologic treatment of resistant hypertension requires the use of effective multidrug regimens. Sequentially combining agents with different mechanisms of actions is intuitive and remains the generally recommended treatment approach. A large number of studies document the added antihypertensive benefit of combining a thiazide diuretic with most other classes of agents. This is particularly relevant in treating resistant hypertension, as studies have suggested that persistent intravascular volume retention commonly underlies resistance to antihypertensive treatment.8 Accordingly, multidrug regimens of 3 or more medications should include a thiazide diuretic unless contraindicated. Hydrochlorothiazide (HCTZ) has been the most commonly used thiazide, in part because it is widely available in fixed-dose combinations. In a rigorous, blinded comparison using 24-hour ambulatory BP monitoring, however, investigators reported that chlorthalidone 25 mg daily provided superior BP reduction compared with HCTZ 50 mg daily in untreated hypertensive patients.9 The greater reduction was true with 24-hour BP monitoring, but even more so with nighttime BPs (a reduction of 13.5±1.9 mm Hg with chlorthalidone compared with 6.4±1.8 mm Hg with HCTZ). This benefit in terms of nocturnal BP control may be particularly relevant as ambulatory nighttime BP levels have been reported to better predict cardiovascular outcomes.10,11 The superior 24-hour antihypertensive efficacy of chlorthalidone undoubtedly reflects in large part the longer half-life of chlorthalidone compared with HCTZ. Thus, in treating resistant hypertension, particularly if the BP remains uncontrolled with use of HCTZ, preferential use of chlorthalidone seems appropriate. In contrast to HCTZ, chlorthalidone is not widely available in fixed-dose combinations, so its use requires prescription of separate pills. Also, chlorthalidone, in our experience, is more likely than HCTZ to cause hypokalemia—requiring closer monitoring of serum potassium levels.
Beyond comparisons of various 2-drug combinations, there are scant data on the relative efficacy of drug regimens consisting of 3 or more drugs. Recommendations are, therefore, intuitive and/or anecdotal. In that regard, it seems most advantageous to combine agents from different classes, as opposed to combining agents with the same or overlapping mechanisms of action. While a few small studies have suggested that the latter strategy may provide substantial additional BP reduction, the studies have generally included only a small number of subjects and have generally not tested maximal doses of the different agents, such that it has not been possible to distinguish a unique additive effect compared to benefit secondary to dose titration.12,13 That is, with continued titration of the initial agent, the same degree of BP reduction may have been observed. We have found a standard triple regimen of a thiazide diuretic, a long-acting calcium channel blocker, and an angiotensin converting enzyme (ACE) inhibitor or an angiotensin receptor blocker (ARB) to be effective and generally well tolerated. In patients with CKD (creatinine clearance <30 mL/min), a loop diuretic may be more effective in reducing volume and BP than a thiazide diuretic. If used, furosemide, because of its short half-life, should be dosed at least twice daily, or alternatively, a longer-lasting agent such as torsemide can be given once daily.
The addition of fourth-, fifth-, and sixth-step medications cannot be standardized and instead must be tailored on an individual basis taking into consideration prior benefit, history of adverse events, contributing factors including concomitant disease processes such as CKD or diabetes, and financial limitations. As discussed below, the use of mineralocorticoid antagonists primarily as fourth-step therapy has been shown to provide significant additional BP reduction that seems to exceed what would be achieved with other available classes of agents, although blinded comparisons confirming such differential benefit are lacking. As opposed to a pure β-blocker, we tend to preferentially use α- β—blockers, as the dual mechanism of action seems to be somewhat more efficacious, although that impression has not been tested in head-to-head comparisons. Centrally acting agents can be effective but have a high incidence of adverse effects, particularly at higher doses. Vasodilators such as minoxidil or hydralazine are also effective, but adverse effects are predictable (tachycardia, fluid retention) and, on rare occasions, potentially dangerous (pericardial effusion, lupus-like syndrome).
USE OF LOW-DOSE ALDOSTERONE ANTAGONISTS
A growing number of studies confirm the antihypertensive efficacy of low-dose aldosterone antagonists as add-on agents in treating resistant hypertension. The strength of these studies is that the degree of antihypertensive benefit has been generally consistent among the studies and appears to exceed what might be expected with use of other classes of agents. The weaknesses of the studies are similar in that they have been generally single-center, open-label assessments. To confirm a unique role of mineralocorticoid receptor antagonism in treating resistant hypertension, blinded, multicenter comparisons must be made with other classes of agents; however, such comparisons will be difficult as they would require attempting to standardize multidrug regimens in a very high-risk cohort.
The effectiveness of aldosterone antagonists in treating resistant hypertension is physiologically consistent with recent studies demonstrating a high prevalence of PA in patients with resistant hypertension. At the University of Alabama at Birmingham, 20% of a series of consecutive patients with resistant hypertension were found to have PA based on suppressed renin activity and high urinary excretion of aldosterone during high dietary sodium ingestion.5 A prevalence of PA of approximately 20% is consistent with results of other investigators in Seattle, WA14 and Oslo, Norway.15 This high prevalence, however, may underestimate the role of aldosterone excess in contributing to treatment resistance, as the persistence of suppressed renin activity in these cohorts exceeded 60% in spite of the use of antihypertensive agents (diuretics, ACE inhibitors, and ARBs) known to stimulate renin release. Such suppressed renin activity has been suggested to represent a variant of aldosterone excess that is not reflected by the measured levels of plasma or urinary aldosterone.16 A broader role of aldosterone in causing treatment resistance may be relevant in explaining the benefit of aldosterone antagonists that exceeds the percentage of patients diagnosed with classical PA.
In one of the earliest reports assessing the benefit of aldosterone antagonists in treating resistant hypertension in patients without evidence of PA, Ramsay et al17 retrospectively compared the antihypertensive benefit of substituting furosemide for a thiazide diuretic or adding spironolactone to a thiazide diuretic in 27 patients with resistant hypertension. The dose of furosemide was 40–80 mg daily, while the dose of spironolactone was 100 mg daily in all subjects except 1, who received 200 mg daily. At 4 weeks' follow-up, furosemide reduced recumbent BP from 185/108 mm Hg to 167/102 mm Hg, while spironolactone reduced BP from 199/111 mm Hg to 169/99 mm Hg. The authors interpreted these results as supporting the role of fluid retention in contributing to resistant hypertension that could be overcome by aldosterone blockade or with the use of more potent diuretics.
Four decades later, French investigators18 prospectively evaluated the effects of adding spironolactone (1 mg/kg) in 23 patients with uncontrolled hypertension in spite of the use of 3 different antihypertensive agents, including a diuretic, in a majority of the patients. After 2 months' follow-up, all subjects had office BP controlled to <140/90 mm Hg, and after 1 month of follow-up the mean 24-hour ambulatory BP had been reduced from 152±2/86±2 mm Hg to 128±2/76±2 mm Hg. The doses of spironolactone used were not reported but presumably would have been approximately 60–70 mg daily. Accordingly, the study suggested that doses of spironolactone less than 100 mg daily provided significant additional BP reduction in treating patients whose BP remained uncontrolled in spite of the use of multiple drugs.
Finding a high occurrence of aldosterone excess in patients with resistant hypertension led us to increasingly use aldosterone blockade as add-on therapy—first in patients with confirmed PA, then in patients with suppressed renin activity but not necessarily high aldosterone levels, and finally in unselected patients whose BP was not responding to multiple medications. We then tested prospectively our positive anecdotal experience, in an open-label evaluation of low-dose spironolactone as additive therapy for treating resistant hypertension.19
In this study, consecutive patients with resistant hypertension were prescribed spironolactone 12.5–50 mg daily and followed for 6 months.19 Patients with congestive heart failure were excluded. A total of 76 subjects were enrolled, including 45 African American subjects. All subjects were evaluated for PA by measurement of plasma aldosterone and renin activity and 24-hour urinary aldosterone excretion. PA was defined as a suppressed renin activity (<1.0 mg/mL/h) and high urinary excretion of aldosterone (>12 μg/24 h) in the setting of high dietary sodium intake (>200 mEq/24 h). All subjects were receiving a thiazide and an ACE inhibitor or ARB.
Spironolactone at a mean dose of 31 mg daily significantly lowered BP at 6 months' follow-up by 25±20/12±12 mm Hg (Figure 1). The BP response was similar in patients with and without PA (Figure 2); patients confirmed to have PA were more likely titrated to spironolactone 50 mg daily (47% vs 12%). Interestingly, the BP response to spironolactone was not predicted by the baseline plasma aldosterone level, plasma renin activity, aldosterone/renin ratio, or 24-hour urinary aldosterone excretion. There was also no racial difference in the BP response, with African American and white patients responding similarly.
Spironolactone at these doses was well tolerated. The most common adverse effect was breast tenderness, which occurred in 10% of the men. Five patients had acute increases in serum creatinine levels (>25% increase or >1.5 mg/dL). In 4 of these subjects, the increase in creatinine occurred in association with a large reduction in BP and resolved on discontinuation of the spironolactone. In 3 patients, the spironolactone was resumed at a lower dose without complication. Hyperkalemia (serum potassium >5.5 mEq/L) occurred in 2 patients. Both of these patients had CKD with a calculated creatinine clearance of <60 mL/min.
Overall, this study suggested that low-dose spironolactone provided significant additional BP reduction when added to multidrug regimens that included a thiazide diuretic and renin—angiotensin system blockers. Spironolactone was well tolerated, although there was risk of hyperkalemia, particularly in patients with CKD. Although patients with PA required higher doses of spironolactone to achieve the same level of BP reduction, the efficacy of spironolactone was not related to measured plasma urinary aldosterone levels. This lack of relation between aldosterone and response to spironolactone suggests that aldosterone excess may play a broader role in causing treatment resistance than is suggested by measured levels of aldosterone.
In an open-label evaluation, investigators in Ireland added spironolactone 50 mg/d to the regimen of patients whose BP remained uncontrolled on 3 or more antihypertensive agents.20 At 14 weeks' follow-up the mean BP reduction was 28±3/13±2 mm Hg. As in the above study, the BP response was not predicted by the baseline aldosterone/renin ratio. Spironolactone was generally well tolerated, although 18.5% of the men developed gynecomastia and 2 patients, both of whom were receiving an ACE inhibitor, developed hyperkalemia.
Amiloride, in blocking the epithelial sodium channel in the distal collecting duct, acts as an indirect aldosterone antagonist. Two studies suggest that amiloride, like spironolactone, provides significant additional BP reduction in treating patients with resistant hypertension. In the first study, Eide et al15 evaluated the BP response to adding amiloride 2.5 mg to the regimen of patients uncontrolled with multiple drugs, including a diuretic. Thirty-eight subjects, all of whom had suppressed renin activity at baseline, were included in the analysis. After 2 weeks of treatment, BP was reduced by 31±31/15±11 mm Hg. In 26 subjects, the amiloride/ diuretic dose was doubled, with an additional reduction in systolic and diastolic BP of 11 and 4 mm Hg, respectively. Amiloride was well tolerated, with a few subjects complaining of abdominal cramps. Serum potassium levels increased overall, but there were no occurrences of hyperkalemia.
Saha et al21 compared the BP effects of amiloride 10 mg daily, spironolactone 25 mg daily, or a combination of both when used as add-on therapy in African American subjects whose BP was uncontrolled on a 2-drug regimen consisting of a diuretic and a calcium channel blocker.21 Ninety-eight subjects were randomized among the 3 treatment groups or placebo with a treatment period of 9 weeks. The mean decreases in systolic and diastolic BP compared with placebo were 12.2±2.2/4.8±1.3 mm Hg for amiloride, 7.3±2.3/3.3±1.4 mm Hg for spironolactone, and 14.1±2.3/5.1±1.4 mm Hg for the combination. Amiloride appeared to be somewhat better than spironolactone in reducing BP; however, amiloride use was associated with significant increases in plasma renin activity, while spironolactone was not. This suggests that the spironolactone as dosed may not have been fully effective and that with titration additional BP benefit might have been observed. Amiloride and spironolactone were both well tolerated, without occurrences of hyperkalemia.
Eplerenone is a selective aldosterone antagonist with a low affinity for progesterone and androgen receptors and, therefore, is less likely than spironolactone to cause breast tenderness, gynecomastia, sexual dysfunction, and menstrual irregularities. Eplerenone has been shown to be an effective antihypertensive agent in treating general hypertensive patients and specifically effective in treating low-renin hypertension.22–24 It has not been evaluated, however, in patients with resistant hypertension. Given its mechanism of action, benefit would be anticipated, but formal evaluation is needed. At this point, eplerenone is a reasonable alternative to spironolactone when the latter is not tolerated. Risk of hyperkalemia is probably the same for eplerenone and spironolactone if used at equipotent doses.
Resistant hypertension is a common medical problem. Successful treatment is predicated on identification and reversal of contributing factors, diagnosis and treatment of secondary causes of hypertension, adoption of appropriate lifestyle modifications, and use of effective multidrug regimens. Inappropriate fluid retention contributes importantly to resistant hypertension; diuretics are essential to maximize a BP response. Recent studies indicate that aldosterone antagonists may provide significant additional BP reduction when added to existing antihypertensive regimens. This benefit is consistent with studies indicating that aldosterone excess commonly underlies resistance to antihypertensive therapy. Aldosterone antagonists are generally well tolerated in this setting, although hyperkalemia can occur, particularly in patients with underlying CKD.
Disclosure. This work was supported by National Heart, Lung, and Blood Institute grants HL075614 and SCCOR P50HL077100.