Pickering clarified the difference between renovascular hypertension and renal artery stenosis:

Most investigators believe that there must be at least a 70% reduction in luminal diameter for a RAS to cause either hypertension or ischemic nephropathy (4,20,21).

Animal models for renovascular hypertension have helped to elucidate the pathophysiology of hypertension in patients with RAS (22,23,24,25). The renin–angiotensin–aldosterone system plays an important role, as shown in Figure 109.1 (26). In animals, the two-kidney, one-clip (2K-1C) model is the classic model for renin-mediated hypertension and is analogous to unilateral RAS in humans. The one-kidney, one-clip (1K-1C) model of renovascular hypertension is a model for volume-mediated hypertension and is analogous to bilateral RAS or RAS to a solitary functioning kidney in humans. Although the acute phases of both of these models are similar, different events occur in the chronic phase.

In the 2K-1C model (unilateral RAS), decreased renal blood flow stimulates the production of renin (Fig. 109.1A). Renin cleaves the proenzyme angiotensinogen to form angiotensin I and in the presence of angiotensin-converting enzyme is converted to angiotensin II. Angiotensin II has several important functions: (a) It elevates blood pressure directly by causing systemic vasoconstriction, (b) it stimulates aldosterone secretion, causing sodium reabsorption and potassium and hydrogen ion secretion in the cortical collecting duct, and (c) it changes the intrarenal hemodynamics, such as diminishing glomerular filtration, by decreasing glomerular capillary surface area and redistributing intrarenal blood flow. The salt and water retained (caused by excess aldosterone production) is rapidly excreted by the contralateral (normal) kidney by pressure natruresis. This produces a cycle of renin-dependent hypertension (23,27).

Administration of an angiotensin-converting-enzyme inhibitor (ACEI) blocks this cycle and, at least early in the course of renovascular hypertension, returns the blood pressure to normal. In the chronic phase the blood pressure remains elevated even though the plasma renin activity returns to baseline. When an ACEI or angiotensin-receptor blocker (ARB) is administered, the blood pressure does not immediately decline, but takes 5 to 7 days (28) to return to normal. The delayed blood pressure response suggests that the change in blood pressure is no longer primarily caused by the direct effects of angiotensin II, but involves some degree of salt retention as well.

In the 1K-1C model of renovascular hypertension, there is a similar decrease in blood flow to the affected kidney(s), acutely causing the secretion of renin and synthesis of angiotensin II and aldosterone (Fig. 109.1B) (26). Angiotensin II directly elevates blood pressure, and aldosterone causes salt and water retention. In this model there is not a normal kidney that can sense the elevated blood pressure, and therefore pressure natruresis does not occur. The increased aldosterone causes sodium and water retention and volume expansion. The expanded plasma volume suppresses plasma renin activity, thus converting the animal from renin-mediated hypertension to volume-mediated hypertension (29). During this stage, administration of an ACEI or ARB does not decrease blood pressure or change renal blood flow (30). Dietary restriction of sodium or administration of diuretics will return the subject to a renin-mediated form of hypertension and restore sensitivity to an ACEI or ARB. Functional renal insufficiency may occur in humans when ACEIs are administered to patients with bilateral RAS or RAS to a solitary kidney especially in the volume-contracted state (Fig. 109.2).

An important clinical concept is that long-standing hypertension may cause changes in intrarenal hemodynamics and structural changes in the microvasculature of the kidney leading to nephrosclerosis (30). Thus, if a patient with
nephrosclerosis later develops unilateral RAS, he or she may behave like a patient with bilateral disease because the contralateral kidney would no longer be able to excrete a salt load normally.

Patients who develop azotemia while receiving angiotensin-converting-enzyme inhibitors or angiotensin II–receptor blocking agents often have bilateral RAS, RAS to a solitary kidney, or decompensated congestive heart failure in the sodium-depleted state (13,21,31,32,33,34).

There are two mechanisms by which renal functional impairment may occur with the use of antihypertensive agents. The first may occur with any antihypertensive agent when a critical perfusion pressure is reached below which the kidney no longer receives adequate perfusion. This has been shown by the infusion of sodium nitroprusside in patients with high-grade bilateral RAS. When the critical perfusion pressure was reached, the urine output, renal blood flow, and glomerular filtration rate declined, and they returned to normal when the blood pressure increased above this critical perfusion pressure (35). The exact pressure necessary to perfuse a kidney with RAS varies with the degree of stenosis and is different among different patients.

The second mechanism is confined to patients receiving ACEI or ARB agents and may occur despite no significant change in blood pressure (32). Patients with severe bilateral RAS or RAS to a solitary kidney may be highly dependent on angiotensin II for glomerular filtration. This is particularly common in patients who receive a combination of ACE inhibitor and diuretic (36) or in patients who are placed on a sodium-restricted diet (37). Under these circumstances, the constrictive effect of angiotensin II on the efferent arteriole allows for the maintenance of normal transglomerular capillary hydraulic pressure, thus allowing glomerular filtration to remain normal in the presence of markedly diminished blood flow (Fig. 109.2). In this instance glomerular filtration is highly dependent on angiotensin II. When an ACEI is administered, the efferent arteriolar tone is no longer maintained and glomerular filtration is therefore decreased. A similar situation occurs in patients with decompensated congestive heart failure who are sodium depleted (34,37).

Shanley suggested that loss of renal mass, loss of nephrons, and renal atrophy predominate in the pathologic picture of patients with renovascular disease (38). There is a point at which many of these changes are reversible, and the goal of management is to identify the patients early and treat them to prevent progressive renal failure. Atheromatous embolization has also been associated with renovascular disease (39). The compilation of injury leads to cortical scarring, interstitial fibrosis, tubular collapse, and destruction of the renal architecture.

Systemic and glomerular pressures play important roles in renal injury in patients with renovascular disease. In the two-kidney, one-clip model of renovascular hypertension, the unclipped kidney demonstrates significantly more vascular and glomerular damage than the clipped kidney (40). It is not clear whether this is due to a direct hemodynamic effect of the elevated blood pressure or to the mitogenic and hypertrophic effects of angiotensin.

More than 90% of all renovascular lesions are secondary to atherosclerosis (4). Although atherosclerotic RAS may occasionally be isolated to the renal artery alone, it is more commonly a manifestation of systemic atherosclerosis involving the aorta and coronary, cerebral, and peripheral vessels. Atherosclerotic RAS most often occurs at the ostium or the proximal 2 cm of the renal artery (Fig. 109.3). Distal arterial or branch involvement is distinctly uncommon (26).

Fibromuscular dysplasia (FMD) is a nonatherosclerotic, noninflammatory disease that most commonly affects the renal arteries and is the second-most-common cause of RAS (41). The most common clinical presentation is that of hypertension in a young woman. Fibromuscular dysplasia has been demonstrated in virtually every vascular bed. Renal artery involvement occurs in 60% to 75% of patients with FMD, followed by extracranial cerebrovascular arteries in 25% to 30%, visceral arteries in 9%, and arteries of the extremities in approximately 5% of patients (41,42). It may present as a systemic disease (affecting a combination of the carotids, mesenteric, subclavian, and/or extremity vessels) in up to 28% of patients (43). The diagnosis is rarely made pathologically but is usually determined by its typical angiographic appearance (Fig. 109.4). Whereas atherosclerosis involves the origin and proximal portion of the renal arteries, FMD characteristically involves the distal two thirds of the artery and may involve the branches (41,44).

The lesions of FMD are thought to be congenital dysplasias with maldevelopment of the fibrous, muscular, and elastic tissues of the renal artery. They are subcategorized according to the layer of the arterial wall involved (45,46). This classification is important because each type of fibrous dysplasia has distinct histologic and angiographic features and each type occurs in a different clinical setting (Table 109.1).

Medial fibromuscular dysplasia has been further divided into medial hyperplasia (the only type in which there is true smooth muscle hyperplasia), perimedial dysplasia, and medial fibroplasia. Medial fibroplasia is the histologic finding in nearly 80% of all cases of FMD. It tends to occur in 25- to 50-year-old women and often involves both renal arteries. It has a “string of beads” appearance angiographically, with the “bead” diameter larger than the proximal, unaffected artery (Fig. 109.4). The areas of stenosis are often overshadowed by contrast medium in the microaneurysms, making the degree of actual stenosis difficult to assess. This beading is due to thickening of the media, interspersed by areas of aneurysmal dilation. Microscopically, the internal elastic membrane is focally variably thinned and lost. Within the alternating thickened areas much of the muscle is replaced by collagen, hence the term medial fibroplasia. In other areas, thinning of the media occurs to the point of complete loss, and microaneurysms can be seen as saccules lined by only
the external elastica. In extreme cases, giant aneurysms may be found in association with medial fibroplasia (Fig. 109.5). Progression to total occlusion is rare in this subtype. Medial fibroplasia responds very well to percutaneous balloon angioplasty. The other types of FMD are beyond the scope of this chapter and are nicely summarized in several recent reviews (41,46).

Other causes of renovascular disease include aneurysm, Takayasu arteritis, atheroemboli, thromboemboli, spontaneous renal artery dissection, arteriovenous malformations or fistulas, trauma (such as from radiation, lithotripsy, direct injury, or surgery), and neurofibromatosis. Retroperitoneal fibrosis producing external compression has also been associated with RAS.

Although there are excellent prevalence data in specific patient populations (e.g., those with coronary artery disease, aortic aneurysm, or peripheral arterial disease), the prevalence of RAS in the general population is not known. Not all cases of RAS are clinically significant. Dustan et al. (51) reviewed 149 aortograms and found that approximately half of patients with 50% or greater RAS did not have hypertension.

In a recent population-based study, Hansen and colleagues reported on the prevalence of renovascular disease in a cohort of elderly patients (52). Eight hundred thirty-four participants of the Cardiovascular Health Study underwent renal duplex ultrasound. Fifty-seven (6.8%) had anatomic renal artery stenosis. There was no difference in the prevalence of RAS in whites (6.9%) and African Americans (6.7%).

Several series have looked at the prevalence of renovascular disease in patients who have atherosclerotic disease elsewhere. To determine the prevalence of atherosclerotic RAS, we studied 395 consecutive patients who had undergone arteriography as part of an evaluation for an abdominal aortic aneurysm, aortoiliac occlusive disease, and peripheral arterial disease. In addition, 78 patients had an aortogram performed for suspected renal artery stenosis (Table 109.2) (1). These patients did not have the usual clinical clues to suggest RAS. High-grade bilateral renal artery disease was present in approximately 13% of patients. In the 319 patients reported in six different studies, 44% had bilateral RAS (9). Other studies showed that 22% to 59% of patients with peripheral arterial disease have significant renal artery stenosis (53).

It has also been established that RAS is common in patients with coronary artery disease. Of 7,758 patients undergoing cardiac catheterization during a 76-month period, 3,987 underwent aortography to screen for RAS at the time of catheterization (54). One hundred ninety-one (4.8%) patients had greater than 75% renal artery stenosis and 0.8% had severe bilateral disease. In the Mayo Clinic series (55), renal arteries were studied at the time of cardiac catheterization in patients with hypertension. Ninety percent of the renal arteries were adequately visualized, and no complications occurred from the aortogram. Greater than 50% RAS was present in 19.2% of patients, greater than 70% in 7%, and bilateral RAS in 3.7%.

Knowledge of the natural history is extremely important in the subsequent management of patients with RAS. Most natural history studies reported in the literature were retrospective studies. The rates of progression ranged from 36% to 71% (56). In Schreiber et al.’s series, only 16% of patients progressed to total occlusion over a mean follow-up of 52 months. However, the rate of progression to total occlusion occurred more frequently (39%) when there was greater than 75% stenosis on the initial renal arteriogram (56).

Zierler et al. (57,58) used renal duplex ultrasound to prospectively study anatomic progression of atherosclerotic renovascular disease. If the renal arteries were normal, only 8% progressed over 36 months. However, at 3 years, 48% of patients progressed from less than 60% stenosis to 60% or greater stenosis. All four renal arteries that progressed to occlusion had 60% or greater stenosis at the initial visit. Progression
of RAS occurred at an average rate of 7%/year for all categories of baseline disease combined.

The effect of RAS on kidney size has been well studied (59,60). Using duplex ultrasound, Caps and colleagues prospectively followed 204 kidneys in 122 patients with known renal artery stenosis for a mean of 33 months. The 2-year cumulative incidence of renal atrophy was 5.5%, 11.7%, and 20.8% in kidneys with a baseline renal artery disease classification of normal, less than 60% stenosis, and 60% or greater stenosis, respectively (p = .009, log rank test) (59).

A question that has really never been satisfactorily answered, however, is how common is atherosclerotic RAS as a cause of end-stage renal disease (ESRD)? Scoble et al. (61) found that atherosclerotic renovascular disease was the cause of ESRD in 14% of patients starting dialysis therapy. Mailloux and colleagues (50) reviewed the causes of ESRD in 683 patients over a 20-year period. Eighty-three patients (12%) had documented RAS as a cause of ESRD. Because these investigators only performed arteriography in patients in whom they highly suspected RAS, it is possible that the true incidence of RAS as a cause of ESRD was seriously underestimated. A recent study reported that 16% of 49 patients starting renal replacement therapy had greater than 50% bilateral RAS or RAS to a single functioning kidney (62). Renal artery stenosis should be searched for in every patient starting dialysis if a clear-cut etiology for the ESRD is not known (63,64,65).

Connolly et al. (48) showed that the 2-year actuarial renal survival (percentage of patients remaining off dialysis) was 97.3% for patients with unilateral RAS, 82.4% for patients with bilateral RAS, and 44.7% for patients with stenosis or occlusion to a solitary functioning kidney. Patients on dialysis have a shortened life expectancy. The average life expectancy in a patient with ESRD older than age 65 years is only 2.7 years (66,67). The survival estimates are even worse if the patient has atherosclerotic renovascular disease as the cause of ESRD. Mailloux and associates (50) showed that median survival for patients with renovascular disease was 25 months compared to 55 months in patients with malignant hypertension and 133 months for patients with polycystic kidney disease. The 2-year survival on dialysis in patients with renovascular disease was 56%, 5-year survival was 18%, and 10-year survival was only 5%. These data underscore the fact that patients with atherosclerotic RAS who progress to ESRD and require dialysis have extremely high mortality rates.

The clinical clues that suggest renal artery disease are shown in Table 109.3 (5,20).

Angiography, the gold standard for arterial imaging, is rarely required to make the diagnosis of RAS. Usually one or more of the noninvasive modalities can accurately make the diagnosis. Exceptions to this general rule may occur in patients with fibromuscular dysplasia that primarily affects the branches of the renal arteries and in patients with renal artery aneurysms (41). Angiography or intraarterial digital subtraction angiography (IA-DSA) has several distinct advantages as an imaging modality. All portions of the renal artery including the intrarenal branches can be well visualized. In addition, accessory renal arteries can be identified and one can determine kidney size and gross function (the presence of a nephrogram). However, angiography is invasive and costly compared to other imaging modalities. Although they are not common, complications associated with angiography include access-site complications such as a hematoma, pseudoaneurysm, or arterial venous fistula, as well as acute renal failure from contrast-induced nephrotoxicity and atheromatous embolization to the kidneys, bowel, or lower extremities. Therefore, angiography is a poor screening test for RAS but an excellent confirmatory test once it has been decided that revascularization is indicated.

Carbon dioxide or gadolinium may be used as a contrast agent in patients with severe renal impairment (78,79,80).

A hotly debated and still controversial issue is the performance of renal angiography at the time of cardiac catheterization. Some cardiologists do it routinely on all patients, whereas others selectively image the renal arteries in patients who have clinical clues that suggest the presence of renal artery stenosis. In the Duke series in which 3,987 patients were screened with abdominal aortography at the time of cardiac catheterization, only 191 (4.8%) of patients had greater than 75% renal artery stenosis and only 33 (0.8%) had severe bilateral disease (54). Similarly, the Mayo Clinic (55) prospectively evaluated 297 hypertensive patients at the time of cardiac catheterization and only 19% had renal artery stenosis greater than 50%, 7% had stenosis greater than 70%, and 3.7% had bilateral disease. They did, however, demonstrate that evaluation of the renal arteries can be completed safely with the addition of only 62 cc of contrast. Therefore, although it is safe to perform angiography at the time of catheterization, the yield is low, and there is some evidence that knowing that there is renal artery stenosis present will tempt the angiographer to stent the renal artery in the absence of clear-cut indications (81,82). Therefore, we are opposed to routine “drive-by angiography” at the time of cardiac catheterization because it adds nothing to the overall management of the patient other than to tempt the angiographer to perform a procedure that is not indicated. However, if the patient has a clear-cut indication for intervention (inability to control the blood pressure with a good antihypertensive regimen, jeopardy of renal function, or recurrent episodes of congestive heart failure) and the clinician is prepared to perform angioplasty and stenting should significant RAS be discovered, then an aortogram at the time of cardiac catheterization should be performed.

Evaluation of the renal arteries is essential in acute or chronic renal failure, poorly controlled hypertension, or recurrent congestive heart failure that is not directly caused by myocardial ischemia. Imaging modalities are also useful for screening for restenosis after percutaneous or surgical revascularization of the renal arteries. Every patient who has undergone percutaneous intervention for RAS should be put in a surveillance program for detecting restenosis (5,77). Technological advances in computed tomographic angiography (CTA) and MRA provide aortic and renal artery imaging that compares well to the quality and accuracy of angiography and have virtually replaced catheter-based angiography for the evaluation of aortic, renal, and peripheral arterial disease (20). Factors such as the patient’s body habitus, the degree of renal impairment, and cost may help to determine which screening test is used. However, institutional experience and expertise are perhaps the most important factors determining which of the noninvasive imaging studies is used as a screening test for renovascular disease.