Distinct hypertensive syndromes causing drug-resistant hypertension are increasingly being recognised. The ability to recognise these disorders has come as a result of a better understanding of their pathogenesis and the availability of sensitive and accurate diagnostic tools. The particular appeal of making the diagnosis centres around the potential curability of the hypertension, with correction of the abnormality through medical and surgical approaches. With the availability of a wide array of specific pharmacotherapeutic agents and the improvements in surgical techniques, excellent clinical outcomes can be obtained and are now more predictable than ever before. This article will deal primarily with the disorders most commonly encountered by clinicians: primary aldosteronism, pheochromocytoma (PHEO) and renovascular hypertension. Emphasis will be on clinical recognition and a rational approach to diagnosis.
Primary aldosteronism is the most common form of secondary hypertension and is often associated with drug-resistant hypertension. The urgency in identifying and treating primary aldosteronism is heightened by data from studies showing that persistent elevations of aldosterone can result in end-organ damage.1–3
The prevalence of primary aldosteronism remains debatable, as studies have been fraught with several limitations, including bias in patient selection and reliance on tests that are not regarded as confirmatory for the diagnosis of primary aldosteronism.4 The actual prevalence of primary aldosteronism among unselected hypertensive patients is estimated to be about 4–5.9 %.5,6 In a retrospective observational study of 1,616 patients with resistant hypertension, Douma et al.7 reported that 11.3 % of the patients had primary aldosteronism based on a confirmatory salt suppression test.
Population at Risk
A high suspicion of primary aldosteronism should be entertained in patients with the following:
- spontaneous or unprovoked hypokalaemia with renal potassium wasting;
- severe diuretic-induced hypokalaemia (potassium ≤3.0 mEq/l) that does not normalise after discontinuation of diuretics for at least four weeks and is unresponsive to angiotensin blockers;
- hypertension with adrenal adenoma;
- resistant hypertension with no other evidence of a secondary cause; and/or
- a family history of primary hyperaldosteronism.
Serum potassium concentration is thought to have poor sensitivity and specificity in the diagnosis of primary aldosteronism.8 However, hypokalaemia, when present, is an important sign of excessive aldosterone production.9 Indeed, hypokalaemia (i.e., serum potassium <3.5 mEq/l), accompanied by renal potassium wasting (i.e., 24-hour urinary potassium ≥30 mEq), was found to have a sensitivity of 0.73 and a specificity of 0.94. When spontaneous hypokalaemia is ≤3.0 mEq/l, the sensitivity and specificity improve to 1.00 and 1.00, respectively. Hypokalaemia ≤3.5 mEq/l provoked by a high salt intake had a sensitivity of 0.86 and a specificity of 0.96.
The ratio of plasma aldosterone concentration (PA) to plasma renin activity (PRA) (PA:PRA ratio) is considered to be the best screening test for primary aldosteronism.8 This ratio is, however, denominator-dependent and cut-off values can vary from 7.2 to 100 ng/dl per μg/ml/hr depending on the assay used for PRA. The ratio also has wide variations in sensitivity (64–100 %) and specificity (87–100 %), with a high false positive rate of 35–50 %.10 A better diagnostic accuracy is obtained when the absolute PA is included as a second criterion in combination with the PA:PRA ratio. The combination of a PA:PRA ratio >30 and PA >20 ng/dl had a sensitivity of 90 % and a specificity of 91 % for aldosterone-producing adenoma.11 At Mayo Clinic, in the US, a PA:PRA ratio ≥20 and PA >15 ng/dl were found in >90 % of patients with surgically confirmed aldosterone-producing adenomas.12
An elevated PA:PRA ratio on screening should be confirmed with salt loading (after correction of hypokalaemia) to assess the suppressibility of aldosterone production. Cleveland Clinic patients in the US ingest a normal diet with oral salt (one teaspoon of table salt) added to food each day for five consecutive days. On the fifth day of increased dietary salt, 24-hour urine is collected for sodium, potassium, aldosterone and creatinine. On the morning after completing the 24-hour urine collection, blood is drawn for a basic metabolic panel, aldosterone and renin activity. A 24-hour urinary aldosterone ≥14 μg (mean + 2 standard deviations above values obtained in essential hypertensives) is definitive evidence of non-suppressible aldosterone production, as long as the 24-hour urinary sodium is ≥200 mEq.13 The development of hypokalaemia with renal potassium wasting provides additional evidence of inappropriate aldosterone production. However, neither non-suppressed PRA nor the absence of hypokalaemia precludes the diagnosis of primary aldosteronism. Drugs that affect aldosterone biosynthesis and/or interfere with renal tubular potassium handling are discontinued four to six weeks before testing.
Under certain circumstances, a confirmatory test is unlikely to add further diagnostic information. For example, a patient with resistant hypertension who presents with serum potassium of 2.6 mEq/l, 24-hour urinary potassium of 60 mEq, plasma aldosterone of 50 ng/dl and PRA <0.65 ng/ml may proceed to an adrenal computed tomography (CT) scan without additional biochemical testing. This patient has severe hypokalaemia with renal potassium wasting, high plasma aldosterone (>20 mg/dl) and a PA:PRA ratio of 61.
The two major causes of primary aldosteronism are bilateral adrenal hyperplasia (BAH), which accounts for 65–70 % of patients, and aldosterone-producing adenoma (APA), accounting for 30–35 % of patients. Unilateral hyperplasia and familial hyperaldosteronism account for 3–4 % of patients with primary aldosteronism. While BAH and APA patients respond well to specific pharmacotherapy, it is important to identify those with APA, as this is a potentially curable condition with surgical intervention.
Adrenal Computed Tomography Scan
As the initial step in subtype differentiation, all patients who are confirmed with primary aldosteronism should undergo an adrenal CT scan. A high-resolution CT scan with contrast and fine cuts (2.5–3.0 mm) is the imaging technique of choice. It is generally more available and less costly than magnetic resonance imaging (MRI). Characteristically, adenomas will be found to be small (<3.0 cm), well circumscribed and homogeneous in appearance, with X-ray attenuation of ≤10 Hounsfield units (HUs) and contrast washout ≥50 % (see Figure 1). The reported sensitivity rates of localising adenoma by CT scan are between 75 and 80 %.14 Nodules that are less than 1.0 cm in size are more likely to be missed by CT scan.
Adrenal Vein Sampling
The Endocrine Society guidelines8 recommend that all patients in whom treatment is practicable and desired should undergo adrenal vein sampling (AVS). There may be some exceptions to this recommendation. For example, a 40-year-old patient presents with resistant hypertension, clear-cut evidence of non-suppressible aldosterone production on suppression testing, and an adrenal CT scan showing a 2.5 cm, well-circumscribed, homogeneous left adrenal mass with HUs <10 and contrast washout >50 %. The contralateral right adrenal gland is completely normal. This patient could proceed to surgery without undergoing AVS, as he or she has clear-cut evidence of an APA on the left side with a completely normal contralateral adrenal gland. In addition, it is unusual for patients <50 years to harbour a non-functioning adrenal adenoma. Another example is that of a 50-year-old patient with resistant hypertension, clear-cut evidence of non-suppressible aldosterone production on testing, and an adrenal CT scan showing bilaterally enlarged adrenal glands with macronodules (see Figure 2). Again, this patient should proceed to medical therapy without AVS, which is unlikely to help.
Interpretation of Adrenal Vein Sampling for Aldosterone
The first step is to determine if both adrenal veins have been successfully catheterised. An adrenal vein:peripheral vein cortisol ratio of at least 3:1 (without adrenocorticotropic hormone [ACTH] stimulation) or 5:1 (with ACTH stimulation) indicates successful bilateral adrenal vein catheterisation with 100 % reproducibility.15 An aldosterone:cortisol (A/C) ratio from the dominant (ipsilateral) adrenal vein divided by the A/C ratio from the non-dominant (contralateral) adrenal vein of ≥4 indicates lateralisation. An A/C ratio from the non-dominant (contralateral) adrenal vein divided by the A/C ratio from the peripheral vein of <1.0 indicates suppression of aldosterone production from the contralateral gland. The sensitivity and specificity rates for AVS are 80 % and 75 %, respectively. An example illustrating AVS results in a patient with a left adrenal mass and biochemical and radiologic findings suggestive of primary aldosteronism is shown in Table 1.
The potential effects of various antihypertensive medications on renin and aldosterone levels should be kept in mind. From a practical standpoint, most drugs can be continued during evaluation.
Beta-blockers lower renin (and this should lead to downstream lowering of aldosterone as well). Angiotensin-converting enzyme (ACE) inhibitors and angiotensin receptor blockers (ARBs) increase renin and decrease aldosterone levels. Direct renin inhibitors (DRIs) decrease both renin activity and aldosterone levels. In patients taking DRIs, ACE inhibitors or ARBs, suppressed renin and increased aldosterone, especially if associated with hypokalaemia, would be highly suggestive of primary aldosteronism. Mineralocorticoid (aldosterone) receptor blockers such as spironolactone and eplerenone should be stopped for at least six weeks, as they lead to elevation in both aldosterone level and renin activity and the ratio can thus be difficult to interpret.
Laparoscopic adrenalectomy is the preferred approach in patients with confirmed unilateral adenoma. Patients with BAH and those who prefer not to undergo surgery can be managed medically. Pharmacological therapy consists of the mineralocorticoid (aldosterone) receptor blockers spironolactone or eplerenone, with or without diuretics.13
PHEO is a rare cause of hypertension but its diagnosis is important. Undetected, it is almost always fatal, whereas the hypertension is usually curable by resection of the tumour. The tumour occurs in one in every 20,000–50,000 patients admitted to hospital. About 1,000 new cases are found annually in the US. PHEO is found in approximately 0.5 % of patients with hypertension.16 It has no sex predilection. It occurs at any age, but it has two peaks of incidence, in childhood and early adulthood. It may become manifest for the first time during pregnancy.
Population at Risk
A high suspicion of PHEO should be entertained in patients with:
- family history of PHEO;
- history of non-cardiogenic pulmonary oedema;
- hypertensive crisis with glucocorticoid or ACTH administration;
- episodes suggestive of acute myocardial infarction but normal coronary artery angiograms;
- enhanced adrenal incidentalomas with HUs ≥22 on CT scan;
- presence of lactic acidosis in the absence of shock; and/or
- hypertensive crisis during surgical procedure.
Clinical and Pathophysiological Correlates
The classic triad of episodic headaches, excessive sweating and tachycardia with cardiac awareness, along with paroxysmal hypertension, may not be seen in all patients with PHEO. Patients with essential hypertension may also present with similar paroxysmal symptoms. Thus the definitive diagnosis of PHEO rests primarily on the demonstration of excessive catecholamine production. If PHEO is responsible for the ‘classic spells’, the biochemical test results are always unequivocally abnormal. Patients with PHEO may be completely asymptomatic and yet may have elevated circulating catecholamines.17 With the widespread use of CT and MRI, approximately 50 % of all PHEOs are initially detected as adrenal incidentalomas in patients without the ‘classic spells’ and frequently without hypertension. It has also been noted that blood pressure has no relationship with plasma catecholamine levels during hypertensive episodes in PHEO.
Biochemical results with a high positive predictive value for PHEO are plasma free metanephrine (MN) >1.21 nmol/l, normetanephrine (NMN) >2.21 nmol/l, plasma catecholamines >2,000 pg/ml, and urinary (MN + NMN) >1.8 mg/24 hours. Biochemical results with a high negative predictive value for PHEO are plasma free MN <0.5 nmol/l, NMN <0.9 nmol/l, plasma catecholamines <1,000 pg/ml, and urinary (MN + NMN) <1.3 mg/24 hours. Biochemical results that are in the indeterminate range are plasma free MN 0.5–1.21 nmol/l, NMN 0.9–2.21 nmol/l, plasma catecholamines 1,000–2,000 pg/ml, and urinary (MN + NMN) 1.3–1.8 mg/24 hours.18–20
When biochemical results are in the indeterminate range, additional evaluation using the clonidine suppression test is useful to differentiate essential hypertension from PHEO.21 Administration of 0.3 mg of oral clonidine suppresses neurogenically mediated catecholamine release (seen in essential hypertension), but does not suppress catecholamine release by tumour cells (which occurs in PHEO).
In clinically suspect patients, the measurement of plasma free fractionated metanephrines provides the best sensitivity (90–95 %).20 However, the test only has a specificity of about 82 % in sporadic PHEO, but an equally high sensitivity in familial PHEO. Measurement of 24-hour urine metanephrines has low sensitivity but higher specificity. Vanillylmandelic acid (VMA) has poor diagnostic sensitivity and specificity, and is not recommended for initial screening.
Medications that can cause false positive elevations of plasma and urinary catecholamines or metanephrines are shown in Table 2. Smoking, caffeine and exercise can transiently increase catecholamine levels; therefore, blood draws for biochemical tests should be done at least one hour after smoking or caffeine intake and after at least 30 minutes of supine rest.
Localising procedures should be performed only after biochemical confirmation of the presence of a PHEO. Adrenal CT scan and MRI have similar sensitivities (95–98 %) and specificities (75–78%) in the localisation of a PHEO. Both have similar positive and negative predictive values (69 and 98 %, respectively).17 On CT scan, the characteristic features of a PHEO is a dense, vascular and heterogeneous appearance with HUs ≥22 and contrast washout <50% (see Figure 3). Metaiodobenzylguanidine (MIBG) scintigraphy is reserved for patients with large tumours (≥5.0 cm in size), patients with multiple adrenal and extra-adrenal sites, patients with focally invasive tumours or metastatic sites, and patients who have definitive biochemical evidence of PHEO in whom other imaging techniques have failed to demonstrate the site of the tumour. In a recent meta-analysis, 123I-MIBG was reported to have a sensitivity of 94 % and a specificity of 92 % in the detection of PHEO.22
Pre-operative Management and Follow-up
Once diagnosed, the management of PHEO is surgical excision (by laparoscopic techniques) carried out by experienced surgeons and anaesthesiologists in a specialised centre. The goals of pre-operative management are twofold: control of hypertension and prevention of post-operative hypotension.
In our clinic, for blood pressure control, we combine a specific postsynaptic α1 receptor antagonist (doxazosin) with a calcium channel blocker (amlodipine). Doxazosin not only inhibits the norepinephrine (NE) effects on postsynaptic α1 receptors in vascular smooth muscle, resulting in vasodilation, but also blocks the NE inhibitory action on insulin release, correcting any existing abnormal glucose regulation. Calcium channel blockers are universal vasodilators that also have cardio- and renoprotective properties. Doxazosin is started at 2 mg at bedtime and increased in 2 mg increments every third day up to a maximum of 10 mg. Depending on the cardiovascular response, amlodipine is then added at 5 mg increments up to a total dose of 10 mg per day. At the same time, the patient is placed on a daily oral intake of least 200–250 mEq sodium. This can be accomplished by adding one measured teaspoon of table salt taken daily (over a 24-hour period) to the usual dietary salt intake. Sitting and standing blood pressures are recorded daily by the patient or during outpatient visits. The target blood pressure is about 130/80 mmHg sitting and a standing blood pressure drop of no more than 20/10 mmHg. Management should obviously be tailored to the individual patient, depending on their co-morbid conditions. A key point in management is that beta-blocking agents should never be started without adequate alpha blockade. Intra-operative hypertensive crises can be managed using intravenously administered sodium nitroprusside, nitroglycerin or nicardipine.
In our clinic, patients are followed up with measurements of fractionated plasma free MN at one, three, six and 12 months post-operatively, and at annual intervals thereafter.
Renovascular hypertension (RVHT) is due to decreased perfusion from renal artery stenosis (RAS), which results in activation of neurohormonal systems. RAS can be due to atherosclerotic disease or fibromuscular dysplasia (usually seen in younger women). Ischaemic nephropathy is difficult to define and is generally used to mean deterioration in renal function attributed to renal artery disease that affects the renal mass. Activation of neurohormonal pathways and oxidative stress contribute to this.
The prevalence of RAS depends on the population that is being screened. The incidence of this disorder rises in patients with severe, refractory hypertension. In patients aged 67 years and older, the incidence of diagnosed RAS was reported to be 3.7 per 1,000 patient-years. An autopsy series reported finding RAS in more than half of 27 % of patients aged over 50 years and in 56.4 % of hypertensive patients. The prevalence was reported to be 10 % in normotensive patients.23,24 RAS is estimated to be the cause of hypertension in only 0.5–4 % of hypertensive patients.25,26
Population at Risk
A high suspicion of RAS should be entertained in patients with a clinical history of:
- recent onset of accelerated or malignant hypertension;
- hypertension refractory to an appropriate three-drug regimen;
- recent loss of blood pressure control;
- recurrent ‘flash’ pulmonary oedema;
- history of tobacco use; and/or
- acute renal failure in patients who are treated with ACE inhibitors or ARBs;
and/or with the following physical and laboratory clues:
- atherosclerotic disease in other vascular beds;
- systolic/diastolic abdominal bruit;
- unilateral small kidney; and/or
- elevated plasma renin activity with spontaneous hypokalaemia.
Diagnostic Approach and Management
Screening for RAS is recommended if an intervention would be planned if clinically significant disease were to be detected. In patients with resistant hypertension who are felt to be at low risk of having RAS, hypertension and other modifiable risk factors should be treated first.
The gold standard test for the diagnosis of RAS is a renal angiogram. Non-invasive options (which are obviously considered first for practical reasons) are listed below.27
- Doppler ultrasound, which has a sensitivity of 0.82 and a specificity of 0.90; Doppler ultrasound is non-invasive and inexpensive, but highly operator-dependent.
- CT angiogram, which has a sensitivity of 0.86 and a specificity of 0.94; it gives excellent image quality for diagnosis, but is expensive, requires use of large contrast load and is associated with radiation exposure.
- Magnetic resonance angiogram, which has a sensitivity and a specificity of 0.88; it gives excellent image quality and is not associated with radiation exposure, but is expensive and has poor quality in the setting of stents and with distal stenosis; additionally, there is a risk of development of nephrogenic systemic fibrosis in the presence of kidney disease.
- Captopril scintigraphy, which has a sensitivity of 0.79 and a specificity of 0.82; it is non-invasive and inexpensive, but less accurate in kidney disease, bilateral disease and obstructive nephropathy.
It is important to keep in mind that essential hypertension often co-exists with clinically silent RAS, and that the presence of stenosis with hypertension does not necessarily equate to having RVHT.
The therapeutic options for RAS are medical management, percutaneous angioplasty with or without stent placement, and surgery. Surgical intervention is considered only in complex lesions. The choice between medical management and percutaneous angioplasty with or without stent placement (‘interventional’ therapy) should be based on a number of factors, including response to medical management, location and significance of the lesion, and patient preference. The possible adverse outcomes of stenting include contrast injury in patients with pre-existing renal impairment, atheroembolism, and stent thrombosis or re-stenosis. The decision to intervene with stenting should be a multidisciplinary one involving both interventionalists and nephrologists.
Stenting has become standard in the endovascular treatment of RAS; however, its role in management (compared with medical therapy) remains a matter of debate. The results of two randomised trials are summarised here; the results of a third trial are awaited. The Stent placement and blood pressure and lipid-lowering for the prevention of progression of renal dysfunction caused by atherosclerotic ostial stenosis of the renal artery (STAR) trial randomised patients to undergo stenting versus medical therapy alone, and the results favoured a conservative approach with focus on cardiovascular risk management, rather than stenting.28 One of the drawbacks of this study was the inclusion of patients with mild RAS, and the trial, with only 140 patients, was underpowered to provide a definitive estimate of efficacy. The Angioplasty and stenting for renal artery lesions (ASTRAL) trial looked at a primary outcome measure of change in renal function over time as assessed by the mean slope of the reciprocal of serum creatinine, and concluded that there was no clear benefit from stenting.29,30 While this study was much larger than STAR (it included 806 patients), one major criticism was the lack of a core central laboratory to interpret imaging studies and the severity of stenosis.
The Cardiovascular outcomes in renal atherosclerotic lesions (CORAL) trial is a large ongoing study examining the effect of stenting versus medical therapy on a primary endpoint of survival free of cardiovascular and renal adverse events (defined as a composite of cardiovascular or renal death, stroke, myocardial infarction, hospitalisation for congestive heart failure, progressive renal insufficiency, or need for permanent renal replacement therapy) in 900 patients.31 Hopefully, it will provide a clearer answer to the question of whether renal artery stenting is beneficial in this patient population.
Currently, intervention with stenting should not be considered in patients whose hypertension can be controlled medically and whose renal function has remained stable over the past 6–12 months; stenosis in itself, even if bilateral, is not an indication for stenting. On the basis of retrospective data, intervention with stenting can be considered in the following groups of patients with a haemodynamically significant stenosis:32–35
- patients with recurrent episodes of congestive heart failure and ‘flash’ pulmonary oedema without an obvious cause, in the setting of bilateral RAS or stenosis to a single functioning kidney;
- patients with rapidly declining kidney function over the past three to six months without an obvious cause, in the setting of bilateral RAS or stenosis to a single functioning kidney; and
- patients with uncontrolled hypertension in spite of optimal, intensive medical management.
Of the secondary causes of hypertension, primary aldosteronism, PHEO and RVHT are the hypertensive syndromes most commonly encountered in clinics. Identifying the population at risk is the initial step in the diagnostic process. An understanding of the pathophysiology of each disorder and familiarity with the available diagnostic tools should lead to accurate and prompt diagnosis. The availability of sensitive and accurate diagnostic tools together with innovative surgical techniques have led to excellent clinical outcomes that are more predictable than ever before.