Resistant hypertension

1. What is resistant hypertension (RH)? Is it the same as refractory hypertension?

Hypertension is considered resistant if the blood pressure (BP) cannot be reduced below target levels ( Box 67.1 ) in patients who are compliant with an optimal triple-drug regimen that includes a diuretic typically with an angiotensin converting enzyme inhibitor (ACEI) or angiotensin receptor blocker (ARB) plus calcium channel blocker (CCB), or those who have controlled BP but are on four or more medications to achieve BP control. RH is increasing in frequency due to increased age and obesity. Most patients with hypertension respond favorably to 1 to 3 antihypertensive drugs. RH is present in about 21% of U.S. adults with hypertension, though a much smaller fraction, about 3%, are truly refractory. Refractory hypertension is an inability to control high BP with the use of five or more classes of antihypertensive agents, including a long-acting thiazide-type diuretic, such as chlorthalidone, and a mineralocorticoid receptor antagonist, such as spironolactone. As with the definition of RH, this diagnosis does not take into account out-of-office BP. Up to 37% of patients labeled as resistant hypertension are actually controlled when BP is measured by 24-hour ambulatory BP monitoring.

Both patients with RH, whether controlled (on four pills) or uncontrolled (on three), and refractory hypertension (uncontrolled on five) have increased cardiovascular risk. This is true for patients with and without chronic kidney disease (CKD). African Americans and patients with CKD are at elevated risk of resistant or refractory hypertension.

Although the definition of RH is currently “in-office BP greater than 140/90,” the SPRINT trial has called into question that BP goal established by clinical practice guidelines, at least in high-risk patients, and the manner of BP measurement. This means that the definition of RH may change too.

If a patient’s BP is uncontrolled on multiple antihypertensive drugs, but one of them is not a diuretic, a diagnosis of RH should be delayed until after a trial of a properly dosed diuretic. It is worthwhile remembering that an “optimal” dose may not necessarily equate to a “full” dose. An optimal dose is the highest dose tolerated by the patient, or a dose governed by concomitant conditions, such as CKD or congestive heart failure.

2. Discuss the epidemiology of RH.

Although not representing typical office practices, the data from the Anti-hypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT) provides an estimate of the prevalence of RH in the community setting. In ALLHAT, 34% of the study participants had BP >140/90 mostly on two medications, with 27% requiring three or more medications. Based on the 5-year observations in ALLHAT, the incidence of RH was approximately 15%. In the Controlled ONset Verapamil INvestigation of Cardiovascular End Points (CONVINCE) study, 18% of the patients had a BP level >140/90 mm Hg on three or more medications.

Patients with diabetes are more resistant to antihypertensive drugs than nondiabetic subjects are, so they will typically require more antihypertensive drugs to achieve goal BP.

3. What are the most common causes of RH?

When a patient with hypertension demonstrates RH, proper management requires the identification of possible etiologies ( Table 67.1 ). Before making drastic therapeutic changes, certain questions should come to the physician’s mind:

• Does the patient truly have “resistant hypertension”?

  • BP measurement is key to the diagnosis of RH. Evaluate BP appropriately with cuffs allowing unobserved automated BP: either the Omron 907 XL, which was used in the SPRINT trial (5-minute rest followed by three readings 1 minute apart), or the BPTru, which measures six readings after 1-minute rest, deletes the first reading, then averages the last five. These methods are accurate, validated, and correlate with target organ damage. On average, using either automated cuff, the systolic BP (SBP) readings are 8 to 15 mm lower than casual office BP readings. There is considerable inter-individual variability in the degree of difference, so one cannot reliably adjust casual office readings to approximate automated readings. The solution is for offices to adopt automated BP machines.

  • In a large Spanish cohort of patients with RH, 24-hour ambulatory BP measurements showed that one-third of patients diagnosed with RH in the office have BPs that were well controlled out of office.

  • The BP cuff needs to circle 80% of the upper arm and needs to be greater than 40% of the length of the upper arm to be accurate. Accurate BP readings require that the patients have their feet on the floor, their back supported, their arm at the level of their heart, and their urinary bladder empty.

• Are there any patient/environmental factors?

  • Nonadherence is common. When blood or urine drug testing is used to confirm adherence of hypertensive patients to their prescribed medication, up to 50% of patients are at least partially nonadherent. The most accurate method to assess adherence, which has not been tested widely in office practice, is high-performance liquid chromatography (HPLC)-mass spectroscopy of blood or urine, to look for prescribed medications. HPLC-mass spectroscopy is an analytical chemistry technique that identifies complex compounds, including antihypertensive medications. Its availability and cost varies. It is widely available in Europe but not in the United States.

  • Alternatives to biochemical testing of compliance include having patients bring pill bottles or calling the pharmacy about refills at each visit. The increase in cardiovascular risk noted in RH has been studied in populations where some nonadherence is very likely. So, although nonadherence may be present in RH, it is not a “cause” per se.

  • Common medications the patient may be taking that will increase BP. Methylphenidate, venlafaxine, and daily use of nonsteroidal antiinflammatory agents can all raise the BP. Vascular endothelial growth factor inhibitors, high-dose steroids, and calcineurin inhibitors all worsen BP. European black licorice, found in complementary alternative medications and chewing tobacco, plus traditional Chinese medicines and other supplements may have enough glycyrrhizic acid to cause pseudohyperaldosteronism. Many supplements do not accurately list their ingredients.

  • Many patients are inadequately diuresed. Hydrochlorothiazide and furosemide do not lower BP over a 24-hour period. Chlorthalidone is a thiazide-type diuretic that is superior to hydrochlorothiazide for controlling BP and preventing cardiovascular events. Its effectiveness is likely due to its long duration of action. Additionally, there are data showing effective BP control down to an estimated glomerular filtration rate of 20, and a larger, confirmatory trial is under way. In patients that require loop diuretics, torsemide is a better choice than furosemide due to its long half-life and predictable bioavailability.

  • Most patients with RH consume a high-sodium diet. The Dietary Approaches to Stopping Hypertension (DASH) diet, along with sodium restriction, can significantly improve BP. In the original study, BPs dropped 8 to 13 mm Hg systolic in the hypertensive sub-cohort. Sustained sodium restriction may result in profound improvements in SBP (13 to 17 mm Hg) in CKD stage 3 and 4 hypertensive patients.

  • Exercise for 30 minutes 5 days a week has been shown to be the equivalent of an additional drug in patients with RH.

  • It has been reported that cigarette smoking can interfere with BP control mechanisms. Smoking cessation should be encouraged to stabilize kidney function and decrease cardiovascular risk.

  • Excessive alcohol consumption (more than 2 oz or 60 mL daily) raises the SBP, sometimes to dangerously high levels.

  • Obesity is an important cause of a lot of hypertension. To date, no careful trials have been performed in RH, but the usual weight-appropriate recommendations for structured weight loss or bariatric surgery (which does reduce BP) may result in improvement in otherwise RH.

  • Pseudopheochromocytoma. Severe emotional stress in childhood or young adulthood, such as incest, witnessed suicide, or chronic severe unremitting stress, may be associated with labile hypertension without overt anxiety. It has been described as a calm panic attack. The preferred drugs include alpha blockers, antidepressants, and β-blockers.

• Does the patient have a secondary form of hypertension, such as primary aldosteronism or renovascular disease?

  • Screen all patients with RH with an aldosterone renin ratio. An elevated aldosterone and suppressed renin is indicative of primary aldosteronism. Fully 20% of resistant hypertensives screen positive for primary aldosteronism. If the screen is positive (aldosterone at or near 15, renin less than 1), follow up with further evaluation. The Endocrine Society updated its guidelines on primary aldosteronism in 2016. If an adenoma is found on a computed tomography scan as part of the evaluation, the patient should also be screened for cortisol excess. Although pheochromocytoma is vanishingly rare, it is easily ruled out with a plasma metanephrine and normetanephrine assay.

  • In older patients, if clinical suspicion is high due to diffuse vascular disease and recurrent bouts of pulmonary edema, or persistent azotemia with the use of an angiotensin-converting enzyme (ACE) inhibitor, further evaluation for renal artery stenosis is indicated, although careful selection of patients for renal revascularization is critical. A subset of patients in the Cardiovascular Outcomes in Renal Atherosclerotic Lesions (CORAL) study without proteinuria had good outcomes with stenting.

  • Consider polycystic ovary syndrome in young women with refractory hypertension. Rare adrenal (11 and 17-hydroxylase deficiency) and mineralocorticoid receptor alterations (Liddles, 11 β-OH SD deficiency), which usually manifests in childhood, may cause RH. It is also important to check thigh BP measurement in patients younger than age 30 to screen for coarctation of the aorta.

• Does the patient have obstructive sleep apnea?

  • Always evaluate patients with RH for sleep apnea. Examine their mouth for a Mallampati score and perform a sleep apnea screen, such as the STOP-BANG or Epworth sleepiness scale. Refer appropriately for a positive screen. Large drops in BP with treatment are disappointingly rare; however, treatment of sleep apnea may improve diabetes and ease weight loss. Patients with excessive daytime sleepiness that improves following continuous positive airway pressure may have a greater drop in BP, which is otherwise typically only 2 to 3 mm Hg.