Author + information
- Received July 18, 2008
- Revision received September 29, 2008
- Accepted October 10, 2008
- Published online March 1, 2009.
- Robert D. Safian, MD, FACC⁎ ( and )
- Ryan D. Madder, MD
- ↵⁎Reprint requests and correspondence:
Dr. Robert D. Safian, Director, Cardiac and Vascular Intervention, Department of Cardiovascular Medicine, William Beaumont Hospital, Heart Center-3rd Floor, 3601 W. Thirteen Mile Road, Royal Oak, Michigan 48073
Renal artery stenosis (RAS) is caused by a heterogenous group of diseases with different pathophysiology, clinical manifestations, treatment approaches, and outcomes. The 2 most common forms of RAS are fibromuscular dysplasia (FMD) and atherosclerosis (ARAS). Renovascular syndromes are broadly classified into renovascular hypertension and ischemic nephropathy, but these terms are misleading, because they imply a causal relationship between RAS, hypertension, and renal dysfunction, which is difficult to prove in humans. Data supporting renal revascularization are limited by heterogeneous causes of hypertension and renal dysfunction, insufficient understanding of the relationship between RAS and nephropathy, inconsistent techniques for revascularization, ambiguous terminology and end points to assess benefit, and lack of large-scale randomized trials. The purpose of this review is to enhance understanding of the epidemiology, clinical markers, and diagnosis of RAS; the relationship between RAS and important disease states; the distinction between renal ischemia and nephropathy; optimal revascularization techniques; and avoidance of renal injury.
Renal artery stenosis (RAS) is caused by a heterogenous group of diseases with different pathophysiology, clinical manifestations, treatment approaches, and outcomes. The 2 most common forms of RAS are fibromuscular dysplasia (FMD) and atherosclerosis (ARAS), whereas inflammatory disease of the arterial circulation and congenital abnormalities are far less common (Fig. 1). Traditionally, renovascular syndromes have been broadly classified into 2 categories: renovascular hypertension and ischemic nephropathy. These categories are potentially misleading, because they imply a causal relationship between RAS and hypertension or renal dysfunction, respectively. Although causal relationships are evident in experimental models of RAS, they are more difficult to prove in human diseases. Furthermore, a causal relationship suggests that revascularization of RAS should favorably impact blood pressure and renal function, yet available clinical data have failed to demonstrate unequivocal benefits of renal revascularization. The purpose of this review is to place RAS in appropriate perspective, particularly with regard to renal revascularization and the importance of renal ischemia and nephropathy. This perspective should incorporate understanding of the epidemiology, clinical markers, and diagnosis of RAS; establish a relationship between RAS and important disease states; distinguish renal ischemia and nephropathy; use optimal revascularization techniques; and avoid renal injury.
Epidemiology of RAS
Fibromuscular dysplasia is an uncommon disease of unknown etiology; typically occurs in women <30 years of age; and often affects the renal, carotid, and femoral arteries. Fibromuscular dysplasia should be considered in young patients if severe hypertension is not associated with obesity, oral contraceptives, or known renal parenchymal disease. Unilateral or bilateral renal FMD might cause renovascular hypertension, but renal failure is unusual (1).
Atherosclerotic renal artery stenosis is a common clinical entity, affecting 7% of patients older than age 65 years and 60% of patients with hypertension, coronary or peripheral artery disease, and renal insufficiency (2). Unlike FMD, ARAS rarely causes renovascular hypertension but is commonly associated with renal dysfunction (3).
Clinical Manifestations of RAS
Hypertension and cardiovascular manifestations
Hypertension manifestations include onset of severe hypertension at age <30 years (FMD) or at age >55 years (ARAS) and resistant, accelerated, or malignant hypertension (3). Cardiovascular manifestations usually occur in the setting of malignant hypertension. The classic manifestation is “flash” pulmonary edema not explained by coronary artery or valvular disease, especially if left ventricular function is normal (4). Other cardiovascular manifestations include severe hypertension associated with acute coronary syndromes, acute aortic syndromes, stroke, transient cerebral ischemia, intracranial hemorrhage, encephalopathy, and papilledema.
Renal ischemia might present as acute renal failure, with a rise in serum creatinine within 14 days of initiation of angiotensin-converting enzyme inhibitors or angiotensin receptor blockers. Although considered a marker for bilateral RAS, this observation is neither sensitive nor specific for RAS (5). Other renal manifestations are subtle or insidious, including unexplained chronic renal failure, small kidney, and asymmetry in renal dimensions (2). Ischemic nephropathy is an important cause of chronic kidney disease and end-stage renal disease, representing the primary etiology of end-stage renal disease in 5% to 15% of patients initiating dialysis each year (6).
Assessment of RAS and Its Clinical Significance
Screening for RAS
There are no guidelines for routine screening for RAS. In some patients, the diagnosis of RAS is made incidentally during angiographic evaluation of lower extremity arterial diseases, whereas in others a high index of suspicion is required, on the basis of existing guidelines (Table 1). The mere presence of angina, congestive heart failure, coronary artery disease, and peripheral artery disease are not strong indications for evaluation of RAS in the absence of other considerations mentioned in the preceding text. Impromptu “drive-by” renal arteriography during unrelated angiographic procedures is not recommended.
Establish the diagnosis of RAS
If clinical manifestations suggest RAS, the contemporary approach is to use renal duplex ultrasound, magnetic resonance angiography (MRA), or computerized tomography angiography (CTA) to identify RAS. Assessment of the renin-angiotensin system is not recommended (2). Invasive angiography is sometimes recommended to confirm the diagnosis of RAS; determine the etiology; identify dual, accessory, or aberrant renal arteries; identify diseases of the abdominal aorta; and evaluate the nephrogram. The angiographic technique is important to minimize renal injury, prevent atheroembolization, and obtain high-quality images. In most cases, abdominal aortography with digital subtraction provides superb images of the abdominal aorta and renal circulation, with a power injector and 10 to 15 cc of contrast (Fig. 2). Because 30% of patients have dual, accessory, or aberrant renal arteries (Fig. 3), selective angiography alone might preclude complete assessment of the renal arteries. Once “anatomic” RAS is recognized, it is important to establish a relationship between RAS and vital organ injury. The complete evaluation of patients with ARAS and vital organ injury must include a baseline assessment of nephropathy and renal ischemia.
Relationship Between RAS and Renal Dysfunction
The renal artery, the kidney, and renal function
In simplistic terms, the kidney is a filter with inflow (renal arteries), outflow (renal veins), and a reservoir (renal pelvis, ureters, and bladder) (Fig. 4). Apart from diseases of outflow (renal vein obstruction) and collection (obstructive uropathy), filter dysfunction might be due to inflow impairment (RAS and renal ischemia) or filter impairment (nephropathy) or both. When RAS and nonvascular etiologies of renal dysfunction co-exist, it might be difficult to establish RAS as the culprit. Patients with nephropathy might not improve after renal artery revascularization, depending on the extent of baseline nephropathy before revascularization and the degree of renal injury after revascularization (2,7). The relationship between renal ischemia and nephropathy is central to understanding published studies and ongoing trials of RAS, and failure to do so is the most important source of ambiguity about the benefits of renal revascularization.
Clinical evaluation of nephropathy
The clinical evaluation for nephropathy includes serum creatinine, urinalysis, renal duplex ultrasound to assess renal dimensions and renal resistive index, and selective renal arteriography in some patients to assess cortical blood flow and intrarenal arteriolar patterns (Table 2,Fig. 5). Individually, none of these parameters is an absolute predictor of outcome, and over-reliance on any single test might exclude patients who might benefit from revascularization (8). In any given patient, certain measures might indicate greater degrees of nephropathy than others, but advanced nephropathy is characterized by proteinuria >1 g/day, renal length <10 cm, and resistive index >0.8 (2).
Serum creatinine is the most common measure of renal function but is limited for assessing the extent of dysfunction or for distinguishing nephropathy from renal ischemia. Serum creatinine is insensitive to glomerular filtration rate (GFR) until 50% to 75% of renal mass has been lost (Fig. 6). Stated in another way, a patient who loses 50% of renal mass (as might occur after nephrectomy or with unilateral renal artery occlusion) should have a normal creatinine; serum creatinine ≥2 mg/dl in a patient with unilateral ARAS is generally indicative of significant nephropathy (9).
Clinical evaluation of renal ischemia
Patients with RAS and abnormal perfusion by objective measurements should be considered to have renal ischemia. Several noninvasive methods have utility for estimating renal blood flow, assessing the hemodynamic significance of RAS, and identifying renal ischemia (Table 3). Nuclear scintigraphy with technetium-labeled pentetic acid (99MTc-DTPA) is reliable for measuring fractional renal blood flow and, when used in conjunction with 125I-Iothalamate, allows accurate measurement of total- and single kidney-GFR (10–12). In patients with unilateral RAS, hypoperfusion of the stenotic kidney is reasonable evidence for renal ischemia; patients with normal renal blood flow might have RAS but not ischemia.
The invasive evaluation of renal ischemia is based on hemodynamic assessment of RAS rather than renal artery perfusion per se. Stenosis severity determined by visual estimates or quantitative angiography has a poor correlation with hemodynamic significance (13). Translesional pressure gradients (TLG) >20 mm Hg, with small catheters or special pressure wires, are considered hemodynamically significant (14). Fractional flow reserve can determine the hemodynamic significance of RAS, and fractional flow reserve <0.80 might predict a favorable blood pressure response to revascularization (13,15). Intravascular ultrasound is extremely useful for assessing vessel dimensions and stenosis severity in FMD patients and, when used with TLG, provides useful assessment of ischemia and improvement after angioplasty. Intravascular ultrasound might be used to guide stenting for ambiguous ARAS (16). Renal frame counts and renal blush scores are used to assess the hemodynamic significance of RAS (17), but high frame counts and low blush scores might be observed with nephropathy without RAS (18–20).
New classification for RAS, renal ischemia, and nephropathy
We propose the following classification to allow identification of patients with and without nephropathy and with and without renal ischemia (Table 4,Fig. 7): Type 1: normal kidneys (no nephropathy); Type 2: nephropathy (parenchymal disease); Type A: no renal ischemia (hemodynamically insignificant RAS); and Type B: renal ischemia (hemodynamically significant RAS).
This classification offers a reasonable framework for evaluation of patients with RAS and allows identification of patients with normal (Type 1) and abnormal (Type 2) kidneys and with normal (Type A) and abnormal (Type B) renal perfusion. The goals of the classification are to provide more consistent terminology, offer a framework for evaluating renal ischemia and nephropathy, and guide management decisions. The best candidates for revascularization are those with vital organ injury, renal ischemia, and no nephropathy.
Renal Artery Revascularization: Technical Considerations
Technique of renal artery revascularization
For FMD patients, balloon angioplasty is the intervention of choice, and stenting is used for bail-out indications. Procedural success approaches 100%, and restenosis occurs in <10% within 10 years (21,22). Renal angioplasty is better in discrete lesions in major renal arteries and worse in diffuse FMD in small segmental, arcuate, and interlobar vessels. Because renal arteriography is not reliable for assessment of FMD stenosis severity or vessel dimensions, we recommend the pressure wire and intravascular ultrasound to assess TLG, vessel dimensions, and stenosis severity. Patients with nonobstructive FMD should be treated conservatively.
In patients with ARAS, stenting is recommended to eliminate elastic recoil, minimize dissection, and maximize lumen enlargement (Fig. 8). Most studies report procedural success rates of 95% to 100%, residual diameter stenosis <10 %, restenosis rates of 10% to 15% within 1 year, and major complications in <2% (14).
There are a number of important technical and procedural considerations to avoid renal artery injury, kidney injury, and atheroembolization. Selective renal arteriography should be guided by abdominal aortography; the catheter-in-catheter or no-touch techniques should be used to minimize contact with the aortic wall and injury to the renal ostium during guiding catheter engagement (23) (Fig. 9). The nephrotoxic effects of radiographic contrast are minimized by maintaining adequate hydration, limiting contrast volume, and using digital subtraction angiography. Renal embolization during revascularization seems to be fairly common (24), and 1 small randomized study suggested potential benefit of a combination of distal embolic protection and intravenous abciximab (25). Because 14% of patients have early renal bifurcations, complete renal “protection” might not be possible. All patients should be evaluated for post-procedural nephropathy and have regular follow-up.
Outcomes After Renal Artery Revascularization
Failure of renal revascularization to cure hypertension
After ARAS revascularization, hypertension cure (normal blood pressure, no medication) is observed in <10% of patients, regardless of the revascularization technique (26,27). The explanations for why renal revascularization does not cure hypertension are somewhat speculative, and available data are limited by heterogeneous causes of hypertension and renal dysfunction, insufficient understanding of the relationship between renal ischemia and nephropathy, inconsistent techniques for revascularization, ambiguous terminology and end points to assess clinical benefit, and the lack of large-scale randomized trials. There is a persistent misperception that ARAS patients have renovascular hypertension, and contemporary reviews continue to use this terminology (14,26). Whereas the experimental Goldblatt models are compelling demonstrations of renin-angiotensin activation due to RAS (28), the mechanisms of hypertension in humans with and without RAS are far more complex and include sympathetic and cerebral nervous system activation, vasoactive oxygen species, abnormalities in endothelial dependent relaxation, and ischemic and hypertensive intra-renal injury (29–31). Patients with ARAS do not have renovascular hypertension, as evidenced by similarities in the extent of renin activation compared with hypertensive patients without RAS and the low cure rate of hypertension after successful revascularization (32,33). The most likely explanations are that patients with ARAS have essential hypertension, many do not have renal ischemia, and unrecognized hypertensive nephropathy leads to self-perpetuating hypertension.
Failure of renal revascularization to improve renal function
Several studies documented improvement in creatinine and in the slope of reciprocal creatinine after stenting, compatible with beneficial effects of revascularization on renal function (34–36). Nevertheless, 25% to 30% have deterioration in renal function despite revascularization. The explanations for failure to improve or stabilize renal function after revascularization are multifactorial, including revascularization of patients without renal ischemia, insensitivity of the creatinine to changes in GFR when <50% of renal mass is revascularized (e.g., unilateral ARAS) (Fig. 6), failure to identify baseline nephropathy, and procedure-induced nephropathy. The key observation in prior studies of ischemic nephropathy is the crucial importance of baseline renal function (36–38): Baseline creatinine >1.5 mg/dl is the single strongest predictor of late death (39), and the risk of renal failure rises 3-fold for each increment of 1.0 mg/dl in baseline creatinine (3).
ARAS and cardiovascular outcomes
Four-year survival rates are 57% and 89% for patients with and without ARAS, respectively, and mortality rates are higher with more severe ARAS and with bilateral ARAS (2,40). Although ARAS adds incremental risk to cardiovascular morbidity and mortality, there are no data that renal revascularization improves cardiovascular outcomes. The CORAL (Cardiovascular Outcomes in Renal Atherosclerotic Lesions) trial is randomizing 1,080 patients with ARAS to optimal medical therapy or to optimal medical therapy plus renal artery stenting (41). Patients must have unilateral or bilateral ARAS, resistant hypertension, and/or chronic kidney disease ≥stage 3. The primary end point is a composite of cardiovascular or renal death, stroke, myocardial infarction, hospital stay for heart failure, progressive renal insufficiency, or the need for renal replacement therapy. The results of CORAL are expected in 2010; a potentially important limitation is the lack of baseline assessment of nephropathy and renal ischemia.
Approach to Specific Clinical Situations
Hypertensive RAS patients without vital organ injury
In older patients with new or refractory hypertension, imaging studies are reasonable to detect ARAS, but revascularization is controversial in the absence of renal ischemia or cardiovascular injury (Fig. 10). For patients with no nephropathy and normal renal blood flow (Type 1A), we would intensify the antihypertensive regimen and follow patients clinically for the development of vital organ injury. For patients with unilateral or bilateral ARAS, no nephropathy, and abnormal perfusion, we would “reclassify” such patients as having renal ischemia (Type 1B).
Hypertensive patients with vital organ injury
Hypertensive patients with manifestations of vital organ injury should undergo an imaging study to detect RAS. The best candidates for revascularization are those with minimal or no nephropathy and renal ischemia (Type 1B). The worst candidates are those with advanced nephropathy (Type 2), especially if renal ischemia is absent (Type 2A).
For asymptomatic patients <30 years old with controlled or resistant hypertension, CTA is reasonable to diagnose FMD. Recommendations for revascularization are influenced by patient age, FMD location and distribution, hemodynamic significance of stenosis, and tolerance of antihypertensive medication. In the majority, angioplasty is appropriate to control hypertension. Because 25% of patients with renal FMD have carotid FMD, carotid duplex ultrasound is recommended if renal FMD is identified. In addition, patients with carotid FMD might have berry aneurysms of the circle of Willis, so intracranial MRA or CTA is advisable, too. Decisions about revascularization of renal FMD are simpler than ARAS, because of the high (>80%) cure rate and durability (90% patency at 10 years) after angioplasty. Because hypertensive FMD patients generally have renovascular hypertension, angiotensin converting enzyme inhibitors and angiotensin receptor blockers are usually effective. Patients who do not respond or develop vital organ injury should be considered for revascularization. Clinical follow-up is recommended; uncontrolled hypertension or new vital organ injury should prompt repeat invasive evaluation.
If studies demonstrate normal kidneys and renal ischemia (Type 1B), we consider this a manifestation of “unilateral” vital organ injury. This form of renal ischemia is not mentioned in American College of Cardiology/American Heart Association (ACC/AHA) guidelines, but such patients could be considered for renal stenting. The physician might be confronted by a more challenging decision in patients with serum creatinine >2 mg/dl and unilateral ARAS. In these patients, nephropathy is highly likely (Type 2), and preservation of renal function is less likely after revascularization. If such patients develop cardiovascular injury, renal revascularization might be reasonable if renal ischemia is present (Type 2B), although renal function might not improve; recommendations in these patients should be individualized.
Bilateral ARAS or ARAS of a solitary kidney
Patients with bilateral ARAS or ARAS of a solitary kidney might have global severe renal ischemia and are more prone to pulmonary edema than those with unilateral RAS. Patients with severe bilateral ARAS, minimal or no nephropathy, renal ischemia (Type 1B), and cardiovascular injury are ideal candidates for renal revascularization. Evidence for nephropathy might be missed if serum creatinine is the only measure of renal function. Such patients should be considered for revascularization if nuclear GFR is <60 cc/min/1.73 m2, even in the absence of cardiac or cerebral dysfunction, before the development of more advanced renal dysfunction. Nuclear blood flow studies and/or invasive assessment of ischemia are useful in patients with bilateral ARAS to identify the more hemodynamically impaired renal artery and to serve as a baseline for follow-up.
Patients with end-stage renal disease
In the absence of diabetes or other confirmed nephropathy, it is reasonable to perform an imaging study to diagnose ARAS in patients who have been on dialysis for <1 year (2). Patients with unilateral stenosis or occlusion are unlikely to benefit, but patients with bilateral ARAS or occlusion might separate from dialysis after renal revascularization (Fig. 11) (42).
Follow-up of patients with RAS
There are no ACC/AHA guidelines for following patients with RAS. Our approach to all RAS patients includes semiannual assessment of blood pressure, serum creatinine, and vital organ injury. We obtain annual or biannual evaluation of nuclear GFR and split renal blood flow for ARAS patients. After stenting for ARAS, we repeat nuclear blood flow studies at 3 months and annually thereafter. Initial improvement in stented-kidney GFR followed by deterioration is suggestive of restenosis.
Despite publication of the 2005 ACC/AHA guidelines, there is persistent controversy about renal artery revascularization. This controversy stems from imprecise understanding of renal vascular syndromes and their relationship to hypertension and renal dysfunction and is compounded by failures to differentiate renal ischemia from nephropathy, resulting in confusing and conflicting data about outcomes. Accordingly, we propose “new rules” for patients with renal vascular diseases (Table 5):
The term “renovascular hypertension” should be reserved specifically for patients with renin-dependent hypertension, in whom revascularization is expected to cure hypertension. For practical purposes, this is true for many patients with FMD but not with ARAS.
Hypertension in patients with ARAS might be classified as controlled, refractory, accelerated, or malignant, depending on the clinical circumstances. Patients with ARAS and hypertension should not be classified as having “renovascular hypertension,” because there is no compelling evidence that ARAS causes hypertension, and cure of hypertension after revascularization is rare.
The term “renal ischemia” should be reserved for patients with RAS and abnormal renal perfusion (unilateral or bilateral).
The term “renal artery stenosis” should be used for patients with “anatomic” stenosis but has no implications regarding renal ischemia.
The term “nephropathy” should be reserved for patients with renal parenchymal disease. “Ischemic nephropathy” should be reserved for patients with renal parenchymal disease associated with longstanding atherosclerosis and intrarenal arteriolar disease. Other forms of nephropathy might be based on the presence of known diseases, such as diabetes (diabetic nephropathy), hypertension (hypertensive nephropathy), or interstitial diseases (interstitial nephropathy) or might be acquired as a complication of revascularization (acute tubular necrosis, contrast nephropathy, and renal embolization).
Medical therapies, particularly antihypertensive drug therapy and therapies to limit ARAS, are the primary therapies for all patients with RAS.
The optimal use of renal artery revascularization is poorly defined. Contemporary decisions about renal revascularization must include an assessment of the severity and functional significance of RAS (renal ischemia), the condition of the kidneys (nephropathy), and the association between RAS and vital organ injury.
The benefits and risks of renal revascularization will be improved by careful patient selection. Better results will be achieved in patients with vital organ injury, renal ischemia, and no nephropathy.
Appropriately designed randomized clinical trials are essential to define the role of renal revascularization. Such trials must incorporate assessment of renal ischemia and nephropathy, because these factors have the strongest influence on outcome.
The authors wish to thank Darcie Brunette for typing and revising the manuscript, Sue Tomaszycki for preparation of figures, and David M. Harvey for artistic design.
- Abbreviations and Acronyms
- American College of Cardiology
- American Heart Association
- atherosclerotic renal artery stenosis
- computerized tomography angiography
- fibromuscular dysplasia
- glomerular filtration rate
- magnetic resonance angiography
- renal artery stenosis
- renal resistive index
- translesional pressure gradient
- technetium-labeled pentetic acid
- Received July 18, 2008.
- Revision received September 29, 2008.
- Accepted October 10, 2008.
- American College of Cardiology Foundation
- Hirsch A.T.,
- Haskal Z.J.,
- Hertzer N.R.,
- et al.
- Zeller T.,
- Frank U.,
- Müller C.,
- et al.
- La Batide-Alanore A.,
- Azizi M.,
- Froissart M.,
- Raynaud A.,
- Plouin P.F.
- White C.W.
- De Bruyne B.,
- Manharan G.,
- Pijls N.H.J.,
- et al.
- Dangas G.,
- Laird J.R. Jr..,
- Mehran R.,
- et al.
- Mahmud E.,
- Smith T.W.R.,
- Palakodeti V.,
- et al.
- Gill W.M.,
- Pudvan W.R.
- Kidney D.,
- Deutsch L.S.
- Tegtmeyer C.J.,
- Elson J.,
- Glass T.A.,
- et al.
- Cooper C.J.,
- Haller S.T.,
- Colyer W.,
- et al.
- Cooper C.J.,
- Murphy T.P.
- Dworkin L.D.,
- Jamerson K.A.
- Goldblatt H.,
- Lynch J.,
- Hanzal R.F.,
- Summerville W.W.
- Johansson M.,
- Elam M.,
- Rundqvist B.,
- Johansson M.,
- et al.
- Watson P.S.,
- Hadjipetrou P.,
- Cox S.V.,
- Piemonte T.C.,
- Eisenhauer A.C.
- Dorros G.,
- Jaff M.,
- Mathiak L.,
- et al.
- Rees C.R.,
- Palmaz J.C.,
- Becker G.J.,
- et al.
- ↵Cardiovascular Outcomes in Renal Atherosclerotic Lesions trial. http://coralclinicaltrial.org. Accessed February 10, 2009.
- Epidemiology of RAS
- Clinical Manifestations of RAS
- Assessment of RAS and Its Clinical Significance
- Relationship Between RAS and Renal Dysfunction
- Renal Artery Revascularization: Technical Considerations
- Outcomes After Renal Artery Revascularization
- Approach to Specific Clinical Situations