Renovascular Hypertension: An Update

Abstract

Renovascular hypertension, the most common remediable cause of elevated blood pressure, is a controversial topic, but most authorities agree on several principles. The absolute risk of renovascular hypertension for a specific patient can be estimated using only clinical information, thereby sparing many patients further expensive and potentially dangerous evaluations. Patients with a high absolute risk of renovascular hypertension should have angiography only if they are willing to undergo revascularization if warranted. A screening test (captopril renography, Doppler ultrasonography, magnetic resonance angiography, or computed tomography) is recommended for those with an intermediate absolute risk. Angioplasty should be offered to patients with fibromuscular dysplasia. Whether intensive medical therapy (including an angiotensin‐converting enzyme inhibitor or angiotensin II receptor blocker) for atherosclerotic renovascular hypertension is improved by angioplasty plus stent placement may be answered by ongoing studies, the largest of which may be the National Institutes of Health‐funded Cardiovascular Outcomes in Renal Atherosclerotic Lesions (CORAL) trial.

In most industrialized countries, renovascular disease is the most common remediable cause of elevated blood pressure (BP), especially among older hypertensive patients. ¹ , ² , ³ , ⁴ , ⁵ , ⁶ , ⁷ , ⁸ Although Goldblatt and colleagues developed the animal model that led to understanding its pathophysiology more than 70 years ago, ⁹ the diagnosis and management of renovascular hypertension have markedly changed over the last 40 years, due to more accurate diagnostic procedures, more specific and effective antihypertensive medications, and the results of randomized clinical trials. ³ , ⁴ , ⁶ , ⁸ This article reviews some of the current controversies regarding risk assessment before screening tests, selective use of angiography, and limitation of invasive but potentially curative procedures to persons who are most likely to benefit.

DEFINITIONS

Renovascular Hypertension vs Renal Artery Stenosis

Unlike most other cardiovascular and renal conditions, renovascular hypertension can be diagnosed only retrospectively. Classically, renovascular hypertension can be correctly and properly diagnosed 6 to 12 weeks after an intervention (see below), only if the BP is lower than it was before the intervention, with the patient taking the same or fewer antihypertensive medications. In contrast to renovascular hypertension (which has a physiologic basis for its diagnosis), renal artery stenosis is an anatomic diagnosis. Classically, renal artery stenosis was diagnosed when there was a >75% narrowing of the diameter of a main renal artery or >50% luminal narrowing with a poststenotic dilatation. These criteria were based on planar images derived from renal angiograms done in the mid‐1970s; typically, a 50% luminal narrowing is the minimum in the current literature.

The distinction between renovascular hypertension and renal artery stenosis has at least 3 important consequences. First, many older persons have relatively advanced renal arterial stenoses on angiography, but few have resistant hypertension. ¹⁰ Second, surgical removal of a small kidney due to presumed ischemic nephropathy (from renovascular hypertension) has been followed by normal BP values in only about 25% of the patients in whom it was attempted in the mid‐1950s. Third, diagnostic performance of screening tests for renovascular hypertension are different from those for renal artery stenosis, because the latter analysis usually includes 2 arteries per person, whereas the former is done based on responses in individual patients.

SUBTYPES OF RENOVASCULAR DISEASE

Fibromuscular Dysplasia

Fibromuscular dysplasia (FMD) is a noninflammatory, nonatherosclerotic vascular disease that preferentially affects small to midsized arteries. ¹¹ Although any vascular bed can be affected, it is most common in the renal arteries (60%–75%, where it preferentially involves the distal two‐thirds of the main renal arteries). The etiology of FMD, the most common cause of renovascular hypertension in young women (15–30 years), is uncertain, but it may be in part genetic. It has been associated with female sex; cigarette smoking; ergotamine, methysergide, and α₁‐antitrypsin deficiency; pheochromocytoma, type IV Ehlers‐Danlos syndrome, Alport's syndrome, cystic medial necrosis, neurofibromatosis, and coarctation of the aorta (the latter 2 especially in children).

FMD is an important subtype of renovascular hypertension for 2 reasons. Unlike atherosclerotic disease, it rarely progresses to renal arterial occlusion and/or ischemic nephropathy. ⁴ , ¹¹ Most important, when found in the main renal arteries, it responds extremely well to angioplasty without stent placement. Most recent series indicate that about 40% to 55% of such patients have their BP normalized, with another 30% to 40% improved after angioplasty. Because FMD occurs most commonly in young women, the prospect of saving years of expensive drug treatment with successful angioplasty is economically attractive.

Atherosclerotic Disease

Probably about 90% of current patients with renovascular hypertension have atherosclerotic disease as the underlying pathologic reason for the arterial stenosis. This progressive, occlusive process typically narrows the ostium and proximal third of the main renal artery, as well as the nearby aorta. As with all other atherosclerotic vascular diseases, it is found with increasing frequency with advancing age and has the usual associated risk factors (diabetes, dyslipidemia, tobacco use, and history of cardiovascular events).

Other (Less Common) Causes of Renovascular Disease

On a population basis, Takayasu's arteritis may be the most important less common cause of renovascular disease, especially in India or Japan. Renal arterial aneurysms are a common finding with medial fibroplasias but are often seen in saccular forms (as large as 2 cm) at the bifurcation of the renal artery. Rarer causes of renovascular disease include nonstenotic, but quite long, aberrant renal arteries, emboli generated during endovascular procedures, aortic dissection, or kidneys that move more than 7.5 cm while changing from supine to erect posture.

ESTIMATES OF PREVALENCE AND RISKS

Traditionally, the prevalence of renovascular hypertension was estimated to be 5% of all hypertensive individuals, but it varied from <1% to >50% depending on the population's characteristics. In a population‐based sample of claims from 1,085,250 Medicare beneficiaries in 1999 through 2001, the incidence of atherosclerotic renovascular disease was 3.7 per 1000 patient‐years. ¹² FMD is more common among young hypertensive women. In the late 1980s, FMD represented 30% to 40% of cases of renovascular hypertension at referral centers, but today it composes <10% of cases, as the general population ages and atherosclerotic disease becomes more prevalent.

The major risks of renovascular hypertension are those associated with persistently increased BP and with atherosclerotic disease, ischemic nephropathy, and a high risk of cardiovascular events. Renovascular hypertension tends to be relatively resistant to usual drug therapies; administration of either an angiotensin‐converting enzyme (ACE) inhibitor or an angiotensin II receptor blocker (ARB) can provoke acute renal failure (characteristic of bilateral disease or stenosis of a solitary kidney). In this setting, recurrent pulmonary edema can be a presenting symptom of renovascular hypertension, and it frequently improves or disappears after opening the artery. Patients with FMD seldom sustain renal artery occlusion or ischemic nephropathy, but this is a major risk for patients with atherosclerotic disease that often leads to end‐stage renal disease. ¹³ In a series of 220 patients with atherosclerotic disease followed by ultrasonography (US) to observe its natural history before cholesterol‐lowering drugs were widely used, progressive renal arterial stenosis was seen in 31% over 3 years, including 18% of originally nonstenotic arteries, with eventual occlusion in 9 of 295 arteries. For patients with <60% stenosis originally, 28% progressed, as compared with 49% in those with stenosis >60%. In addition, progressive renal cortical atrophy was noted in 21% of patients when the original degree of renal arterial stenosis was >60%. ¹⁴ In the Cardiovascular Health Study, renal artery stenosis (detected by duplex US in 6% of persons aged ≥65 years) was associated with a 1.96‐fold increased risk of coronary events, independent of baseline blood pressure. ¹⁵ Since atherosclerosis is a systemic disease, it is hardly surprising that individuals with renal artery lesions have a higher risk of cardiovascular events. ¹⁶ In one study during a 2‐year period, cardiovascular events in patients with newly diagnosed atherosclerotic renovascular hypertension were about 4 times more common than in the general population and were more than 10 times more common than adverse renal events, including end‐stage renal disease. ¹²

PATHOPHYSIOLOGY

Classically, the pathophysiology of renovascular hypertension involves progressive stenosis of the renal artery, leading to hypoperfusion of the juxtaglomerular apparatus, increased release of renin, and increased production of angiotensin II. This leads to increases in sympathetic nerve activity, intrarenal prostaglandin synthesis, and aldosterone synthesis and decreased nitric oxide production; most important for the development of hypertension, a direct decrease in renal sodium excretion results. This sequence has been validated acutely, originally in Goldblatt's dogs, but the situation in chronic renovascular disease is more complicated. Over time, the increased plasma renin activity decreases as plasma volume expands, especially when chronic kidney disease is present. During the chronic phase, both BP and intravascular volume can be reduced by angiotensin II antagonists or by removal of the arterial stenosis.

In some animal models, removal of the arterial stenosis does not result in an abrupt or complete decrease in BP, compared with that of age‐matched control animals. This phenomenon may be important in humans, as revascularization surgery has been more successful in lowering BP if hypertension has been present for <5 years before the operation.

DIAGNOSTIC EVALUATION

Although the algorithm for evaluation of suspected renovascular hypertension is controversial, the Figure outlines a common approach. An initial estimation of the absolute risk of renovascular hypertension can be based solely on clinical clues. Identifying an abdominal bruit is said to be the most cost‐effective of these. This is often a high‐pitched, holosystolic bruit with a short diastolic component—unlike the bruit often heard in older persons with an atherosclerotic aorta, which is usually a short rough systolic murmur. If the patient is unwilling to accept surgery (should it be required, even to repair a dissection or perforation during angioplasty), medical management alone is advised. Renal angiography can be performed in individuals at high risk for renovascular hypertension. A sensitive screening test can be offered to those with an intermediate probability of renovascular hypertension. The result will dichotomize those who do not need further testing (but only medical management) and those who should have another test (including angiography). The cutoffs for the latter 2 steps and which screening test should be recommended for most patients is debatable. ⁵ , ¹⁷

Figure.

A diagnostic algorithm for renovascular hypertension. RAS indicates renal artery stenosis; ACE, angiotensin‐converting enzyme.

INITIAL RISK ESTIMATION

Many of the characteristics that distinguish individuals with renovascular compared with primary hypertension (Table I) were identified during the 1970s in a study of 2442 hypertensive patients, of whom 880 had renovascular disease (35% of whom had FMD). Dutch investigators have advanced a “clinical prediction rule” that they derived from one cohort of patients, validated in a separate cohort, and revalidated in a third cohort of patients with drug‐resistant hypertension (35 with and 145 without renovascular hypertension). ¹⁸ A nomogram provides the prior probability of renovascular disease, from a sum of the number of “points” assigned to aspects of the individual's history, physical examination, and simple laboratory studies. This method does not incorporate the BP or serum creatinine response to a renin‐angiotensin blocker.

Table I.

Clinical Clues to Renovascular Hypertension

Characteristic	Approximate Relative Risk (vs Primary Hypertension)
Abdominal bruit (high‐pitched holosystolic with diastolic component)	5.0
Recent loss of BP control (or onset of hypertension)	2.0
Unilateral small kidney	2.0
Keith‐Wagener‐Barker grade III or IV optic fundi	2.0
History of accelerated/malignant hypertension	2.0
Unprovoked hypokalemia (potassium level <3.4 mEq/L)	2.0
Increase in serum creatinine level after ACE inhibitor or ARB therapy	1.8
No family history of hypertension	1.8
Atherosclerotic disease in another vascular bed	1.8
Elevated plasma renin activity	1.8
History of cigarette smoking	1.7
Recurrent pulmonary edema	1.5
Proteinuria	1.4
Older age (per decade of life)	1.2
Hypertension refractory to an appropriate 3‐drug regimen	1.2
Abbreviations: ACE, angiotensin‐converting enzyme; ARB, angiotensin II receptor blocker; BP, blood pressure.

SCREENING TESTS

Several reasonably accurate screening tests for renovascular hypertension have been developed. Some are based on physiologic parameters (eg, renin activity or blood flow to each kidney), some are more anatomically based (magnetic resonance angiography [MRA], computed tomographic angiography [CTA]), and some combine aspects of each (Doppler US, captopril scintigraphy [CS]). The sensitivity and specificity of several of the older tests (eg, plasma renin activity, renal vein renin determinations, technetium‐99^m diethylenetriamine pentaacetic acid [Tc^99m‐DTPA] renal scan) are improved after acute administration of an ACE inhibitor.

In 2001, Dutch investigators reported a selective overview of the world's literature on the performance of screening tests for renal artery stenosis. This end point was chosen because many studies (especially in MRA and CTA) report only relationships between the arterial appearance in the screening test and angiography (ie, analysis per artery), rather than per patient (as would be the case if renovascular hypertension were the end point). They concluded that CTA and gadolinium‐enhanced MRA were the best tests. ¹⁹ Three years later, however, their prospective, multicenter, comparative study of these 2 modalities (vs angiography) in 356 patients indicated that neither could be recommended. ²⁰ An updated (1990–2006) summary of a more inclusive literature review of the sensitivity and specificity of each of these screening tests for renal artery stenosis is given in Table II and is discussed below. ²¹

Table II.

Performance Characteristics (Weighted Averages of The World's Literature 1990–2006) and Advantages/Disadvantages of 4 Screening Tests for Renal Artery Stenosis

Screening Test	Captopril Scintigraphy	Doppler Ultrasonography	Magnetic Resonance Angiography	Computed Tomographic Angiography
No. of publications	69	47	30	14
No. of patients/arteries	5282	2653	1623	1225
Sensitivity (range)	0.79 (0.27–1.00)	0.83 (0.17–1.00)	0.88 (0.54–1.00)	0.86 (0.63–1.00)
Specificity (range)	0.82 (0.44–1.00)	0.82 (0.63–1.00)	0.88 (0.23–1.00)	0.94 (0.63–1.00)
Advantages	Noninvasive, not expensive, may predict BP results after revascularization	Noninvasive, inexpensive, may predict BP results after revascularization	No iodinated contrast needed; excellent image quality	Excellent image quality
Disadvantages	Less accurate in renal impairment, bilateral disease, obstructive uropathy	Operator‐dependent; less useful in obesity, bowel gas, branch lesions, FMD	Expensive; poor images with stents or distal stenoses (eg, FMD); overcalls moderate stenoses; risk of gadolinium‐associated fibrosing dermopathy	Expensive, time‐consuming to process and interpret; not widely available; large amount of contrast sometimes needed
Abbreviations: BP, blood pressure; FMD, fibromuscular dysplasia.

Captopril Scintigraphy

Several agents (Tc^99m‐DTPA, ¹²¹I‐hippurate, or technetium‐99^m mercaptoacetyltriglycine [Tc^99m‐MAG₃, also known as Tc^99m‐mertiatide]) can image the kidneys after an oral dose of captopril. Although CS is now widely available, relatively inexpensive, and simple to perform and consensus criteria for its interpretation exist, ²² the results of CS are very heterogenous (P<10⁻⁸ by Riley‐Day test) across the world's literature. Some of the variability may be due to different isotopes (eg, MAG₃ is better for detecting bilateral disease), unusual characteristics of included patients (decreased accuracy may exist in blacks and individuals who take calcium antagonists), or different diagnostic criteria across studies. In addition, bilateral renal arterial disease, obstructive uropathy, and an elevated serum creatinine level (>2.0 mg/dL) all reduce the accuracy of CS using Tc^99m‐DTPA, the most commonly used isotope. Using patient‐/artery‐weighted averages, this test is about 79% sensitive (range, 27%–100%) and 82% specific (range, 44%–100%) for detecting renal artery stenosis. Several retrospective analyses suggest that CS may predict BP outcomes after revascularization. In hypertensive patients with normal renal function, CS had an overall sensitivity and specificity of about 90% each for renovascular hypertension; the mean positive predictive value in 291 patients from 10 studies was 92%. ²³ Unfortunately, the only prospective randomized clinical trial to date (discussed below) showed no relationship between CS and BP response following angioplasty. ²⁴ For these and other reasons, CS is no longer recommended as a screening test for renal artery stenosis in the 2005 American College of Cardiology/American Heart Association guidelines. ⁴

Doppler US

Duplex US provides both anatomic and physiologic information by directly identifying renal arteries (using B‐mode US) and providing hemodynamic measurements within them (Doppler flow studies). However, Doppler US has even more variable performance characteristics than CS (P<10⁻¹⁴ by Riley‐Day test). Some of its disadvantages (time‐consuming, operator‐dependent, poor‐quality images due to obesity or overlying bowel gas) can be overcome by scanning patients in the fasted state, after a bowel preparation similar to that undertaken before colonoscopy. Most reports of Doppler US use 50% stenosis as the lower limit of detection, since distinguishing between 50% to 69% and ≥70% stenosis is often difficult. Doppler US has about 83% sensitivity (range, 17%–100%) and 90% specificity (range, 63%–100%) for renal artery stenosis. The most important feature of successful Doppler US may be the renal resistive index. In several (but not all) series, values <80 mm Hg have predicted improved BP following revascularization. ²⁵ This observation has not yet been validated in a randomized clinical trial. A decision analysis (based on 74 patients in Quebec) suggested that Doppler US is more cost‐efficient but less sensitive than MRA. ²⁶

Magnetic Resonance Angiography

Twenty‐nine studies have compared MRA, typically done today with phase contrast and/or gadolinium enhancement, with renal angiography (Table II). In the 2001 review of Vasbinder and associates, ¹⁹ gadolinium‐enhanced MRA and CTA had identical and nearly perfect performance characteristics, but the prospective study reported by the same group in 2004 indicated that both interobserver variation and sensitivity were much poorer than previously noted. ²⁰ Their enrolled population had a high prevalence of FMD (36%), which typically has stenoses in the distal two‐thirds of the renal artery, an area not well visualized by the MRA technique. Across all studies, MRA had a sensitivity of about 88% (range, 54%–100%), a specificity of about 88% (range, 23%–100%), and no significant inhomogeneity (P=.12). MRA was more accurate than CS or US (each vs angiography) in 41 patients with a 75% prevalence of renal artery stenosis, ²⁷ as well as in 58 patients with a 77% prevalence of disease (by transstenotic pressure gradient) compared with CS, US, and CTA. ²⁸

Most of the technical challenges related to image acquisition (eg, duration of breath holding), contrast injection, and patient positioning have now been overcome. Its major advantages include excellent image quality, utility in patients with advanced renal impairment, and no need for potentially nephrotoxic radiocontrast. Nephrogenic fibrosing dermopathy is a very rare but serious consequence of gadolinium infusions; some institutions prohibit gadolinium administration without dialysis for patients with stage 3 or higher chronic kidney disease. Other limitations of MRA in the renal bed include its expense; claustrophobia (said to affect about 10% of patients); a tendency to overestimate moderate (ie, 40%–69%) stenoses; the need for carefully timed intravenous injection of gadolinium; reduced accuracy in small, branch, and distal renal arteries; attenuation of signal by indwelling stents; and lack of functional information in the results.

So far, only a single study has compared outcomes after angioplasty based on MRA measurements before the procedure. ²⁹ Despite unusual criteria for a “successful angioplasty” (reduction in diastolic BP >15% and/or reduction in serum creatinine >20%) and stratification of the results based on “normal renal volume” and a calculated “renal flow index,” the authors reported that MRA had a 91% sensitivity and a 67% specificity for predicting outcomes in their 23 patients. This experience awaits replication by others. In a recent cost‐utility analysis, MRA led to the lowest direct costs of any primary screening test if the prevalence of renal artery stenosis in the tested population was >20%, although essentially any strategy involving successful revascularization saved more lives than did medical treatment alone. ³⁰

Computed Tomographic Angiography

The newest common screening test is CTA, which requires intravenous radiocontrast dye and more time and effort in reconstructing 3‐dimensional images of the renal arteries. CTA had nearly perfect performance characteristics in the original analysis of Vasbinder and associates. ¹⁹ Even with the disappointing results in their prospective study and much better performance in subsequent reports, its overall sensitivity is about 86% (range, 63%–100%), with a specificity of about 94% (range, 63%–100%). Only about 24% of the significant inhomogeneity (P<.0001) across the 14 reported studies could be attributed to the prospective study of Vasbinder and associates ²⁰ ; nearly all of the remainder is due to 4 studies that report nearly perfect correlation with renal angiography.

CTA is not yet as widely available as the other screening tests, and so far it has not been correlated with BP‐lowering outcomes after intervention. It requires a good deal of computer technology and programming expertise to identify all possible non‐planar arterial segments, so the time to reconstruct and interpret images is longer than other screening tests. The volume of intravenous contrast that is necessary to obtain good images is a concern. Like MRA, CTA is not quite as accurate in small, branch, or distal renal arteries, and claustrophobic patients are uncomfortable. Indwelling stents are not as big a concern as they are with MRA. Cost‐effectiveness calculations have not yet been done for CTA, but it is likely that they should be similar (if a bit more expensive) than those done for MRA. A “urine CT attenuation ratio” (left/right density, in Hounsfield units, of proximal ureters or renal pelvis obtained 4–5 minutes after injection of contrast) >1.22 had a specificity and sensitivity of 95% and 96%, respectively, in 28 patients with angiographically proven renal artery stenosis, as compared with 48 controls. ³¹

Angiography

The gold standard for the diagnosis of renal artery stenosis has long been renal angiography. Unfortunately, it provides no functional information about renal excretory function and carries the risks of anaphylactoid shock, radiocontrastinduced renal failure, and complications related to vascular access. Individuals with a very high absolute risk (typically ≥70%) of renovascular hypertension should proceed directly to renal angiography, rather than have an imperfect screening test, because the clinician would likely regard any “normal” screening test as a false‐negative and recommend angiography anyway.

Intravenous digital subtraction renal angiography was developed in the early 1980s to avoid arterial puncture and the risks associated with it. Although this technique cannot be followed as quickly by angioplasty (because the contrast is delivered very quickly into a peripheral vein), the delay can allow for discussion as to whether revascularization is truly appropriate (see below). Today, few measure pressure gradients across apparent stenoses found after intra‐aortic renal angiography; instead, intravascular ultrasonographic probes (especially for FMD ³² ) or (more commonly) guide wires precede angioplasty catheters. A recent study of 17 patients showed better prediction with renal fractional flow reserve (measured after papaverine injection) than translesional pressure gradients. ³³

One of the major controversies in renovascular hypertension today is the expanding role of cardiologists in the “fortuitous” diagnosis and subsequent immediate, catheter‐based management of renal artery stenosis. ⁷ , ¹⁰ , ¹⁷ , ³⁴ Much of the 3.9‐fold increase in Medicare claims for renal artery revascularization between 1996 and 2000 has been attributed to cardiologists doing “drive‐by” renal arterial injections after coronary angiography. So far, only one group has stratified patients undergoing cardiac catheterization by criteria that might put them at risk for renovascular hypertension. ³⁵ These included (1) severe atherosclerosis, (2) severe or resistant hypertension, (3) unexplained renal impairment, and (4) history of acute pulmonary edema. They found that 39% of their 837 patients had renal atherosclerosis, with 14% having stenosis ≥50% and 7% having stenosis ≥70%. The latter group was more likely to be older and female and to have poorer renal function, higher BP levels, and carotid arterial disease. Although this report carefully avoids mention of whether any patient underwent renal angioplasty, the authors intend to follow these patients for cardiovascular and renal complications, which will add to our knowledge of the natural (or postintervention) history of patients with “fortuitously discovered” renal artery stenosis, which may have a different prognosis than disease discovered after a premeditated search.

Although indications for renal angiography during and after cardiac catheterization have been proposed, ¹⁰ many have questioned whether it should be done. Two reports suggest little need for the procedure. In a series of 68 patients with renal artery stenosis found during aortography done for other reasons at the Mayo Clinic, there was no change in BP, an increase in antihypertensive medications from 1.6 to 1.9 per patient, and only a slight increase in serum creatinine (1.4 to 2.0 mg/dL) during 39 months of medical management alone. ³⁶ In a multicenter British experience involving 85 patients with incidentally discovered renal artery stenosis, few required renal revascularization, and 24 of the 27 deaths during 2 years of follow‐up were unrelated to the renal arteries. ³⁷ The real hazards associated with stent placement (atheroemboli, dissections, thrombosis, and renal failure) must be weighed against the very controversial but potential benefits on BP, progressive renal disease, and/or recurrent pulmonary edema. ⁴ , ⁷ , ¹⁰ , ³⁴ , ³⁸ It is not yet clear that renal artery stenosis discovered “fortuitously” has the same response to therapy as one found after an intensive, premeditated search based on prior probabilities.

THERAPY

Fibromuscular Dysplasia

Currently, angioplasty (alone) is the treatment of choice for fibromuscular dysplasia. Technical success and 1‐year restenosis rates vary between 83% and 100% and 5% and 11%, respectively. ³ , ¹¹ In 11 large series, 77% of 506 patients who had primary angioplasty had either normalization or improvement in BP. ¹¹ , ³⁹ A shorter duration of hypertension, younger age, and a lesion in a main renal artery predict success. ¹¹ In a 7‐year follow‐up, 34% of patients required repeat angioplasties, sometimes years after an originally successful procedure. ³⁹

Atherosclerotic Renal Artery Stenosis

Three major treatment options exist: surgical revascularization, angioplasty with or without stenting, or medical management. A recent systematic review of these concluded, ³⁸ “The evidence from direct comparisons of interventions is sparse and inadequate to draw robust conclusions.” Some factors that can assist decision making with an individual patient are given in Table III. Whenever medical management is ineffective in controlling BP or a progressive decline in renal function occurs, revascularization can be more strongly considered. ⁴⁰

Table III.

Factors Influencing Selection of Patients for Revascularization

Favorable Response After Revascularization
Recurrent “flash” pulmonary edema
Renal resistive index <80 mm Hg by Doppler ultrasonography
Progressive, ongoing decline in renal function
Recent institution of dialysis in a patient with suspected ischemic nephropathy
Acute, reversible increase in serum creatinine level after angiotensin‐converting enzyme inhibitor or angiotensin II receptor blocker
Refractory hypertension despite an appropriate 3‐drug regimen
No Favorable Response After Revascularization
Blood pressure <140/90 mm Hg on <3 antihypertensive drugs
Normal renal function
Renal resistive index ≥80 mm Hg by Doppler ultrasonography
History or clinical evidence of cholesterol embolization
Heavy proteinuria (>1 g/d)
>10 Years' history of hypertension
Unilateral small kidney (<7.5 cm in length)
Renal artery stenosis <70%

Surgical Revascularization

Surgical revascularization for renal artery stenosis is currently less popular than angioplasty with or without stent placement. However, many surgeons assert that angioplasty may either only postpone the “definitive procedure,” or, worse, require an emergency operation if the angioplasty goes awry. It is therefore useful and customary to arrange “surgical backup” when scheduling renal angiography with possible angioplasty. Several experienced surgical teams have reported an 80% to 90% rate of cured or improved BP after an operation, but some used “softer” end points (eg, a reduction in the number of antihypertensive pills, but no change in BP). Patients with normal renal function fare better than those with reduced kidney function. A recent series of 247 patients reported similar outcomes whether open surgical or percutaneous revascularization was performed. ⁴¹ Traditional aortorenal bypass and renal endarterectomy are currently less popular than bypasses from nonaortic donor (splenic, celiac, mesenteric, or hepatic arterial) sites. These newer and more elaborate procedures limit manipulation of the diseased aorta and minimize atheroembolism at the expense of a somewhat higher perioperative mortality rate (2%–6%). Most of the deaths are related to graft failure or other complications of widespread atherosclerotic vascular disease.

Angioplasty

Angioplasty (without stent placement) for atherosclerotic disease has been less successful than for FMD. In a review of large series of adults undergoing renal angioplasty through 1995, only 65% of 1664 procedures resulted in normalized or improved BP. About 19% were technical failures, and the restenosis rate at 1 year was about 13%. Less success occurred after angioplasty of ostial lesions, sequential stenoses of a single artery, or stenoses in multiple renal arteries to the same side. ⁶

The results of uncontrolled series reporting renal angioplasty to preserve renal function are difficult to interpret, due to lack of a comparable control group. Overall, most large series report little change in serum creatinine concentration or other measures of renal function after angioplasty (compared with preprocedure), although this may well be due to about the same proportion of patients who experience deteriorating renal function as those who improve.

Angioplasty Plus Stent Placement

The addition of an expandable stent to balloon angioplasty in the renal arterial bed has many theoretical advantages, particularly in locations at high risk for restenosis. Stent placement reduces or resolves complications due to local dissection, prevents elastic recoil (thought to be involved in acute restenosis and thrombosis) and nearly eliminates pressure gradients across lesions after angioplasty. Because of these advantages, many operators now prefer to place a stent whenever possible in the renal arterial bed, despite the paucity of published comparative data and the relatively short follow‐up available since its introduction. In a randomized trial of angioplasty plus stent placement vs angioplasty alone in 84 patients with ostial lesions, restenosis was much less common after 6 months in those receiving stents (25% vs 70%), but there were no differences across groups in BP or deterioration in renal function. About a third of the patients had major complications related to the procedure. ⁴²

The largest experience with renal artery stent placement comes from a registry of 1058 patients followed for at least 6 months. ⁴³ Technical success was universal; overall, there was a significant decline in BP (168±29/84±15 to 147±21/78±12 mm Hg; P<.05), a slight decrease in antihypertensive medications per patient (2.4 to 2.0), and improved renal function (serum creatinine changed from 1.7±1.1 to 1.3±0.8 mg/dL) during a 4‐year follow‐up, during which the overall mortality was 26%, which validates the high mortality risk in persons with a documented atherosclerotic burden. Other smaller and more recent reports show similar results, although about the same number of patients experience deteriorating renal function as improve after the procedure. Restenosis rates vary between 10% and 30% (depending on length of follow‐up). Stent placement may have special benefits in renovascular hypertension with deteriorating (as opposed to abnormal but stable) renal function.

Medical Management

Intensive therapy to control all atherosclerotic risk factors (eg, antihypertensive drugs, aspirin, 3‐hydroxy‐3‐methylglutaryl coenzyme A reductase inhibitors, smoking cessation, and glycemic control) can be universally recommended for all patients with renovascular disease. Without therapy, the long‐term prognosis is at least as poor for these patients as for diabetics and patients with a previous myocardial infarction. ¹⁶

The major concern about intensified antihypertensive drug therapy is the risk of acute deterioration in renal function that sometimes occurs when an ACE inhibitor or ARB is added to the regimen. These drugs are effective in reducing BP in 86% to 92% of patients with renovascular hypertension (most commonly in combination with a diuretic and a calcium antagonist) and in the earliest published experience, require discontinuation in only about 5% during the first 3 months. The increase in serum creatinine values typically reverts to baseline after stopping the ACE inhibitor or ARB; better BP control, despite an acute deterioration in renal function after treatment with an ACE inhibitor or ARB, can be an indication for renal revascularization.

CLINICAL TRIALS AND META‐ANALYSES THEREOF

Several clinical trials related to therapy of renovascular hypertension have been organized, but crossovers, other confounders, and the continuing evolution of what is considered state‐of‐the‐art treatment protocols complicate their interpretation. The latter problem especially affects early trials involving angioplasty without stent placement, which has gradually fallen into disfavor, and more recent studies in which differential use of ACE inhibitors or ARBs occurred across randomized arms. ³⁸ The most recent systematic review of management strategies found no “good” quality studies, and only 4 (3 of which were cohort studies) with “high” applicability. ³⁸ The absence of solid clinical trial data comparing angioplasty plus stent placement with maximal medical therapy was one of the major reasons for the Stent for Atherosclerotic Ostial Stenosis of the Renal Artery (STAR) study ⁴⁴ in Holland, governmental funding of the Angioplasty and Stent for Renal Artery Lesions (ASTRAL) trial in Great Britain, ⁴⁵ and the Cardiovascular Outcomes With Renal Artery Lesions (CORAL) trial in the United States. ⁴⁶

The 3 randomized trials comparing angioplasty with medical therapy for renal artery stenosis are summarized in Table IV. The best known and most widely quoted of these was the Dutch Renal Artery Stenosis Intervention Cooperative (DRASTIC) trial ²⁴ ; many of its features have generated controversy. The categoric BP responses at 12 months are often overlooked: the angioplasty group had more improved (68% vs 38%), fewer worsened (9% vs 33%; P=.002), and more normalized (7% vs 0%). Within the group originally assigned to drug therapy, those who later underwent angioplasty had a significantly greater decrease in BP from 3 to 12 months than those who were maintained on drug therapy alone (P<.0001), suggesting a problem with the intent‐to‐treat analysis. According to expert radiologists who reviewed all angiograms after randomization, 5 patients in each group had stenosis <50% (and did not meet one of the original inclusion criteria). Renal angiography was repeated at 12 months in 91 of the original 106 patients; 23 of 48 (48%) in the angioplasty group had restenosis ≥50%, compared with 35 of 43 (81%) in the drug treatment group (including 4 occlusions); these restenosis rates are much higher than seen in previous studies. A full 35% of the patients had “normal” CS findings, which had been reported to predict little BP response to angioplasty; this may be the reason that baseline CS results were not predictive in this cohort. Two meta‐analyses of these 3 trials of angioplasty compared with medical therapy have come to slightly different conclusions, despite considering the same 210 patients. ⁴⁷ , ⁴⁸ In one, no significant differences were found in any end point (BP, medications, or renal function). ⁴⁷ Patient‐specific data were available to Ives and associates, ⁴⁸ who reported a slightly larger overall reduction in BP in the angioplasty group (6.3/3.3 mm Hg; P=.02/.03) vs drug treatment. The change in serum creatinine values was not quite significant (P=.06), but favored the angioplasty group.

Table IV.

Summary of Trials Comparing Angioplasty With Medical Therapy for Renal Artery Stenosis

First Author	Plouin 49	Webster 50	Van Jaarsveld 24
Intervention (No. of patients)	Medical Rx: 26 (7 with later angioplasty); 23 angioplasties (2 stents; 3 with later angioplasty)	Medical Rx: 30 (5 with later angioplasty); 21 angioplasties; 2 nephrectomies; 2 bypasses	Medical Rx: 50 (22 with later angioplasty); 56 angioplasties (unsuccessful in 4; 2 with stent; 3 with later surgery)
Primary end point	Ambulatory BP after 6 months	Changes in BP and serum creatinine at 6 months vs baseline	Office B P, serum creatinine, creatinine clearance after 12 months
No differences in:	Primary end point, clinic BP measured by oscillometric device	Primary end points, cardiovascular outcomes (3–54 months)	Primary end points
Significant differences in:	Fewer BP medications and physician‐measured office BP in angioplasty group	Lower BP in those with bilateral disease who received intervention; 40 complications after 135 angioplasties	Fewer BP medications in angioplasty group at 3 months; see text for further discussion
Abbreviations: BP, blood pressure; Rx, therapy.

Three comparative randomized trials have recently been launched in Holland, the United Kingdom, and the United States (Table V). Each of these has short‐term end points (BP and renal function), but will observe all patients for cardiovascular and renal events during several years of follow‐up. The hope is that these trials will have few crossovers and other confounders so that a more “evidence‐based” recommendation can be given about angioplasty plus stent placement, in addition to intensive medical therapy, with identification of independent predictors of the success and failure of each strategy.

Table V.

Summary of 3 Ongoing Clinical Trials of Medical Therapy ± Angioplasty With Stent Placement in Patients With Atherosclerotic Renal Artery Stenosis

	STAR	ASTRAL	CORAL
No. of patients	140	750–1000	1080
Inclusion criteria	Ostial renal artery stenosis ≥50%, CrCl <80 mL/min/1.73 m2	Renal artery suitable for angioplasty ± stent placement; no prior revascularization procedure for atherosclerotic renovascular disease	80%–99% renal artery stenosis, or 60%–80% stenosis with ≥20 mm Hg trans‐stenotic gradient; systolic BP ≥155 mm Hg on ≥2 BP meds
Interventions	Angioplasty + stent vs medical Rx alone	Angioplasty ± stent vs medical Rx alone	Angioplasty + stent vs medical Rx alone
ACE inhibitors and/or ARBs	Only as a “last resort,” after randomization	Not recommended postrandomization	Highly recommended postrandomization
Primary end point	20% Reduction in CrCl	Mean slope of 1/Scr vs time	Composite of CV death, stroke, MI, HF hospitalization, or ESRD
Secondary end points	Acute complications, late complications, renal artery occlusion, doubling of Scr, ESRD, BP, pulmonary edema; CV morbidity/mortality	BP, urinary protein excretion, serious vascular events, ESRD, angiography patency at ≅12 months	Mortality, subgroup analyses, 1/Scr vs time, BP, renal artery patency, renal resistive index, quality of life, cost‐effectiveness
Recruitment period	≅1 Year (?)	6 Years	≅4 Years
Planned mean follow‐up	2 Years	1 Year	2+ Years
Abbreviations: ACE, angiotensin converting‐enzyme; ARBs, angiotensin II receptor blockers; ASTRAL, Angioplasty and Stent for Renal Artery Lesions; BP, blood pressure; CORAL, Cardiovascular Outcomes in Renal Atherosclerotic Lesions; CrCl, creatinine clearance; C V, cardiovascular; ESRD, end‐stage renal disease; HF, heart failure; MI, myocardial infarction; Rx, therapy; Scr, serum creatinine; STAR, Stent for Atherosclerotic Ostial Stenosis of the Renal Artery Study.

CONCLUSIONS

Although there are many unanswered questions about renovascular hypertension, most authorities agree that: 1) the absolute risk for the disease can be estimated with reasonable accuracy, using only clinical information, thereby sparing many patients further evaluation; 2) patients with a very high absolute risk of disease should proceed to angiography if they are willing to undergo revascularization; 3) a screening test should be done in those with an intermediate absolute risk of disease, and the choice of test may depend more on local expertise and cost than on a comparison of published performance characteristics; 4) angioplasty should be offered to patients with FMD; and 5) the question of whether intensive medical therapy (including an ACE inhibitor or ARB) for atherosclerotic renovascular hypertension is improved by angioplasty plus stent placement may be answered by ongoing research.