CHAPTER 106 Renal Artery Hypertension

The classic studies by Goldblatt in the 1930s have shown that renal artery stenosis (RAS) is the underlying cause of renovascular hypertension (RVH).¹ Although only 5% of cases of hypertension can be attributed to RAS, it is a potentially curable cause of hypertension. In most cases, RVH is caused by atherosclerotic RAS or fibromuscular dysplasia (FMD).² Atherosclerosis accounts for 70% to 90% of cases of RAS and usually involves the ostium and proximal third of the main renal artery.²^,³ Fibromuscular dysplasia (FMD) frequently involves the distal two thirds of the renal artery and its branches. It is most often characterized by aneurysmal dilations interspersed with localized narrowings presenting as a so-called string of beads or string-of-pearls appearance on angiography.⁴

Together, atherosclerosis and FMD constitute primary renal artery disease. Conversely, there are many diseases with secondary involvement of the renal arteries causing small vessel and intrarenal vascular disease. The latter category of diseases, also called secondary renal artery disease is, however, beyond the scope of this chapter.

Intra-arterial digital subtraction angiography (IA-DSA) is traditionally regarded as the definitive test to diagnose the presence of RAS. However, both the invasive nature of IA-DSA and the difficulty in assessing the pathophysiologic significance of stenotic lesions have encouraged the search for widely available noninvasive or minimally invasive diagnostic tests. In addition, IA-DSA is not a perfect test for the detection of RAS because it is subject to substantial interobserver variation.⁵

With the introduction of contrast-enhanced magnetic resonance angiography (CE-MRA), multidetector computed tomography angiography (MDCTA), and color Doppler ultrasonography, alternatives for IA-DSA have emerged. In this chapter, an overview of these techniques is provided, as well as a discussion of their relative merits and shortcomings.

RENAL ARTERY STENOSIS

Atherosclerotic Renal Artery Disease

Atherosclerosis accounts for the vast majority of cases of RAS and usually involves the ostium or proximal renal artery. This form of RAS is more prevalent in patients with advanced atherosclerosis, such as older patients and patients with diabetes mellitus, coronary artery disease, and/or peripheral arterial occlusive disease.

Fibromuscular Dysplasia

FMD is a relatively uncommon cause of RAS in comparison to atherosclerosis. Principal forms of FMD include intimal fibroplasia (5% of cases), medial hyperplasia (uncommon), medial fibrodysplasia (85% of cases), and perimedial dysplasia (10% of cases). The cause of FMD remains unknown at present. It is thought that about 10% to 30% of cases of RAS are caused by FMD. It had been thought that FMD is more often encountered in young women between 15 and 50 years of age, but because it is not a lethal disease, older women may also have the disease.

PREVALENCE AND EPIDEMIOLOGY

Atherosclerotic Renal Artery Disease

The prevalence of RAS in the general population is low. Among patients diagnosed with hypertension, RAS is present in 1% to 6%.⁶ On the other hand, atherosclerotic RAS is a common and progressive disease in patients with preexisting manifestations of atherosclerosis or clinical clues suggestive of RAS (see later, “Imaging Indications and Algorithm”). In selected cohorts of patients, such as those undergoing coronary or peripheral angiography, the prevalence of atherosclerotic RAS may increase to more than 20%. Among patients with atherosclerotic RAS, progressive stenosis has been reported in 51% of renal arteries 5 years after diagnosis (including 18% of initially normal vessels). Only 3% to 16% of renal arteries became totally occluded and renal atrophy of more than 60% developed in 21% of patients with RAS.²

Fibromuscular Dysplasia

FMD is responsible for 10% to 30% of cases of RAS.²^–⁴ In one of the largest studies on imaging of RAS, the Dutch Renal Artery Diagnostic Study in Systolic Hypertension (RADISH),⁷ the prevalence of FMD, as shown by IA-DSA, was 27 of 356 patients (7.6%), all of whom were referred for imaging of the renal arteries because of therapy-resistant hypertension and/or clinical clues suggestive of RAS, including a baseline blood pressure higher than 95 mm Hg. In this study, FMD comprised 38% of all cases of RAS.

ETIOLOGY AND PATHOPHYSIOLOGY

The relationships between renal artery stenosis, hypertension, and renal excretory disfunction are not straightforward but are complex. Renal artery stenosis may occur as an isolated finding or in combination with hypertension, renal insufficiency, or both. The underlying pathophysiologic mechanisms are incompletely elucidated. Current understanding ²^,⁸ is that a decrease in renal perfusion pressure activates the renin-angiotensin system, which leads to the release of renin and the production of angiotensin II. This, in turn, has a direct effect on sodium excretion, sympathetic nerve activity, intrarenal prostaglandin concentrations, and nitric oxide production. When all these factors are present, the end result is renovascular hypertension. When hypertension is sustained, plasma renin levels may decrease, partially explaining the limitations of renin measurements for identifying patients with renovascular hypertension; this phenomenon is referred to as reverse tachyphylaxis.

MANIFESTATIONS OF DISEASE

Clinical Presentation

RAS is one of the most common secondary causes of hypertension in the adult population, especially among those with other risk factors for the development of atherosclerosis. Laboratory studies are neither sensitive nor specific for the diagnosis of RAS. However, laboratory measurement of renal function should be performed to exclude other potential secondary causes of hypertension.

In addition to hypertension, patients may present with ischemic nephropathy. Ischemic nephropathy is defined as an obstruction of renal blood flow that leads to ischemia and excretory dysfunction.⁹ The pathophysiology of ischemic nephropathy is not fully understood but it is present when the serum creatinine level is elevated because of loss of renal mass as a consequence of long-standing hypertension. In general, loss of more than 50% of renal mass is usually associated with elevated creatinine levels. It is estimated that ischemic nephropathy is present in up to 24% of patients with unexplained chronic or progressive renal failure.

A more acute presentation of the renal artery stenosis is acute, or flash, pulmonary edema. RAS should be suspected in such patients, especially when there are signs of uncontrolled hypertension or ischemic nephropathy or other clinical clues suggestive of RAS.

Imaging Indications and Algorithm

The rational use of noninvasive imaging techniques to detect RAS is important to maximize the diagnostic performance of the test. It is well known in epidemiology that positive and negative predictive values depend on the prevalence of disease. A general rule of thumb is that diagnostic tests perform better when the prevalence of disease is higher. Therefore, clinical clues are used to increase the a priori chances of a patient having RAS (Fig. 106-1). Clinical clues suggestive of RAS include the following: are the presence of diastolic blood pressure of 95 mm Hg or higher; an epigastric, subcostal, or flank bruit; sudden, accelerated, or malignant hypertension; unilateral small kidney; hypertension in children or young adults (prior to age 35) or adults older than 55 years; hypertension and unexplained impairment of renal function; sudden worsening of renal function in a patient with known hypertension; hypertension refractory to three or more appropriate antihypertensive drugs; impairment of renal function in response to angiotensin-converting enzyme inhibitor; and the presence of extensive occlusive disease in the coronary, cerebral, and/or peripheral circulation.³ Using these clinical clues, the prior probability of RAS can be increased to approximately 20% or higher.⁷^,¹⁰

FIGURE 106-1 Algorithm for evaluating patients in whom renal artery stenosis is suspected. Clinical follow-up includes periodic reassessment with duplex ultrasonography, magnetic resonance angiography, and nuclear imaging to estimate fractional blood flow to each kidney. The treatment of risk factors includes smoking cessation and the use of aspirin, lipid-lowering agents, and antihypertensive therapy.

(From Safian RD, Textor SC. Renal-artery stenosis. N Engl J Med 2001; 344:431-442.)

There are many noninvasive methods used to assess the presence of RAS. Currently used modalities are duplex and Doppler ultrasonography (DUS), magnetic resonance angiography (MRA), and CT angiography (CTA), as well as nuclear medicine tests such as captopril renography. At present, however, there is no consensus on the best noninvasive method for screening for renal artery hypertension. Captopril renal scintigraphy was historically a very popular study that has been replaced at many sites by Doppler ultrasound because of its lack of ionizing radiation exposure and broad availability. In the late 1990s and early 2000s, there was much excitement about MRA and CTA, but the first large-scale truly blinded trials yielded disappointing sensitivities and specificities for the detection of RAS. In retrospect, this was likely the result of a suboptimal choice of imaging parameters—mainly, spatial resolution that was too low. Today, this is no longer an issue because over the past 7 or 8 years spatial resolution has improved dramatically for CTA and MRA techniques, and there is nearly universal consensus that these methods can reliably detect and rule out the presence of RAS. Therefore, invasive catheter angiography remains the gold standard for renal artery imaging but is relegated primarily for therapeutic intervention or diagnosis in cases in which noninvasive studies are not definitive. In clinical practice, individual sites have developed their own imaging algorithms based on clinical access to scanning equipment, available expertise, and/or a specific population.

Apart from determining which technique is best for the detection of RAS, there is an ongoing and controversial discussion about the value of imaging techniques. It has recently been shown in several large trials that invasive treatment of RAS by endovascular or surgical means, rather than medical therapy, does not improve renal function in many patients (see later). The future challenge of imaging in RAS is therefore not only to detect RAS reliably, but also to identify patients for whom invasive intervention might still be of added value over medical therapy alone.

Imaging Techniques and Findings

Once a patient has been identified as being at high risk for RAS the choice of the best diagnostic test remains controversial.

Radiography

Apart from incidental discovery of calcified atherosclerotic plaque formation in the perirenal aorta or renal arteries, there is no role for radiography in the work-up of suspected RAS.

Ultrasound

Ultrasonography has traditionally been used for imaging renal parenchyma, detection of nephrolithiasis, and hydronephrosis. In the past decade, with the advent of color Doppler techniques, ultrasonography has been applied with increasing success to diagnose RAS.

Ultrasonographic examination of the renal arteries is entirely noninvasive because it does not require injection of contrast medium. In contrast to CE-MRA and MDCTA, however, it does require fasting for 8 hours prior to the examination.¹¹

RAS can be detected by color DUS using peak systolic velocity in a stenosis or by poststenotic flow phenomena. Most researchers have used 180 to 200 cm/sec at the point of maximum stenosis as the cutoff point to establish 50% or higher RAS ¹² (Fig. 106-2). However, because of obesity, excessive bowel gas, or inability of the patient to sustain a breath-hold, about 10% to 20% of the examinations are unsuccessful. In addition, a study by Nchimi and colleagues¹³ has demonstrated that the accuracy of DUS for the detection of accessory renal arteries is still low. An indirect approach to assessing the severity of RAS is by measuring acceleration in intrarenal arteries distal to a stenosis and the presence of a tardus-parvus waveform. The tardus-parvus phenomenon is used to describe a change in the Doppler blood flow velocity waveform that may be found downstream from significant arterial stenoses. Tardus refers to delayed or prolonged early systolic acceleration (more than 0.07 second¹⁴), and parvus refers to diminished amplitude and rounding of the systolic peak.¹⁵ By combining direct and indirect measurements, the diagnostic yield may be increased.

FIGURE 106-2 Color duplex ultrasonography (printed in gray scale) in a 69-year-old female patient with unexplained hypertension and suspected left renal artery stenosis. The peak systolic velocity in the left renal artery was 124.0 cm/sec (A), indicating a normal renal artery. Velocities measured in the left renal cortex (B) and medulla (C) also indicate normal renal perfusion.

The most widely used ultrasonographic parameter to assess the functional significance of RAS is the resistance index (RI). The RI can be calculated from a Doppler spectrum, and is defined by the equation

Velocities are measured by obtaining Doppler signals from segmental arteries, typically in the upper, middle, and lower portions of each kidney (Fig. 106-3). When measuring the RI, it is important to keep in mind that pulse rates below 50 or above 70 beats/min, or measuring during inspiration or a Valsalva maneuver, may lead to less accurate measurements. Radermacher and associates¹⁶ have shown that in patients with 50% or greater stenosis in at least one renal artery, RI values above 0.80 are highly sensitive and specific to identify patients in whom angioplasty or surgery will not improve renal function, blood pressure, or kidney survival. However, a potential source of bias in this study was that only in patients in which ultrasound detected 50% or greater RAS was therapy considered. In addition, it is not known whether elevated RI values alone are predictive of therapeutic failure. The same group showed that an RI value above 0.80 measured at least 3 months after transplantation is associated with poor subsequent allograft performance and death.¹⁷