Applied Radiology

Kishore J. Harjai, MD, FACC is an Interventional Cardiology Fellow at the Division of Cardiology, William Beaumont Hospital, Royal Oak, MI. He received his MD from Seth G.S. Medical College, Bombay, India.

Atherosclerotic renal artery stenosis (ARAS) is common, yet is frequently overlooked, in patients with resistant hypertension, diabetes mellitus, and coronary or peripheral vascular disease. In patients with ARAS manifesting as resistant hypertension or renal insufficiency, percutaneous revascularization should be strongly considered to improve blood pressure control and renal function. Percutaneous stent graft repair of abdominal aortic aneurysm (AAA) is a relatively new technique indicated for patients who are at high risk for AAA rupture but are not candidates for conventional surgery. The long-term safety of this technique needs to be demonstrated before it is widely accepted for routine clinical use.

Atherosclerotic renal artery stenosis (ARAS) is a frequently overlooked progressive disease, which is sometimes associated with refractory hypertension and renal insufficiency. Although rare in the general adult population, the incidence of ARAS is much higher in patients with hypertension, diabetes, renal insufficiency, or evidence of atherosclerosis elsewhere ^1,2 (Table 1).

Anatomy and pathophysiology

Atherosclerotic renal artery stenosis typically involves the ostial and proximal portions of the renal arteries. In advanced cases, bilateral ARAS may be present, and the segmental and intrarenal portions of the renal arteries may be involved with consequent loss of renal mass. Atherosclerotic renal artery stenosis is associated frequently with coexisting nephropathies (eg, hypertensive nephrosclerosis or diabetic disease).

Renovascular hypertension (RVH), seen in some patients with ARAS, is related to increased renin secretion by the affected kidney in response to low hydrostatic pressure in the afferent arterioles, the location of the juxtaglomerular renin-secreting granular cells. ³ Increased renin secretion leads to higher production of angiotensin II (a potent vasoconstrictor) and aldosterone (which promotes salt and water retention). In patients with unilateral ARAS, the normal kidney maintains natriuresis and prevents volume retention; this compensation is lost in patients with bilateral renal artery stenosis (RAS) or solitary kidney ARAS. Thus, RVH is driven by vasoconstriction in unilateral ARAS and by vasoconstriction as well as volume retention in bilateral or solitary kidney ARAS. Renal insufficiency as a manifestation of ARAS occurs in patients with bilateral or solitary kidney ARAS and is related to loss of renal mass from chronic ischemia.

Clinical manifestations

Most patients with ARAS are asymptomatic. Atherosclerotic renal artery stenosis may be diagnosed in a significant minority of patients with coronary artery disease if screening abdominal aortography is performed. In an unselected population of patients undergoing cardiac catheterization, 30% had some ARAS and 15% had ARAS with lesions >50% severity. Age, severity of coronary artery disease, congestive heart failure, female gender, peripheral vascular disease, hypertension, renal insufficiency, and cerebrovascular disease are independent correlates of ARAS. ^4,5 Clinical markers of significant ARAS are shown in Table 2. Patients with these features should be screened aggressively for the presence of ARAS.

Diagnosis

The noninvasive tests most frequently used in the diagnosis of ARAS are shown in Table 3. In a meta-analysis of 55 studies involving patients with suspected RVH, Vasbinder and colleagues ⁶ estimated the accuracy of computed tomographic angiography (CTA), magnetic resonance angiography (MRA), ultrasonography, captopril renal scintigraphy, and post-captopril renin assay for the diagnosis of ARAS. They found that accuracy varied greatly for all tests. Receiver-operating characteristic curves showed that computed tomographic angiography and gadolinium-enhanced MRA were superior to the other tests. ⁶ Ultrasonography and captopril renal scintigraphy had similar diagnostic performances, and both were superior to the post-captopril renin assay. Comparative evaluation of these tests is limited because some measure anatomic severity (eg, CTA and MRA), whereas others assess the functional significance of ARAS. The accuracy of these tests is validated against an anatomic gold-standard test, intra-arterial renal angiography. Results of functional noninvasive tests may be judged as false-negative if they do not reveal a moderate, but hemodynamically insignificant, lesion that is found on angiography. Conversely, functional tests may be erroneously regarded as false-positive if a lesion of borderline angiographic severity is hemodynamically significant.

Other functional tests that have been used in the past to diagnose renal artery stenosis (RAS) are plasma renin activity and renal vein renin ratio. The results of these tests are confounded by use of concurrent medications and do not predict response or cure of hypertension.

Angiography is currently the gold-standard method for the definitive diagnosis of RAS. It can be performed as an outpatient procedure with 4F, 5F, or 6F catheters, with excellent visualization of the renal arteries and a very low complication rate. Compared with selective renal angiography, abdominal aortography has the advantages of visualizing accessory renal arteries and the abdominal aorta with a single injection, but may require more contrast and can miss ostial renal artery lesions.

Natural history of ARAS

Untreated RAS is a progressive disease. In 41 patients with RAS and medically controlled blood pressure selected for nonsurgical management, Dean et al ⁷ documented worsening of renal function in 46% and decreased renal length in 37%. Zierler and colleagues ⁸ monitored the progression of 139 renal arteries in 80 patients using Duplex ultrasound. At 2-year follow-up, none of the arteries that were normal at baseline showed progression of disease. Of 35 arteries with <60% stenosis at baseline, 42% progressed to >60% stenosis; of the 63 arteries with >60% stenosis at baseline, total occlusion occurred in 11% and was associated with marked reduction in kidney length. In 1189 patients who underwent abdominal aortography as part of diagnostic cardiac catheterization study, Crowley et al ⁵ showed significant angiographic progression of RAS in 11% of patients at a mean follow-up of 2.6 years.

The role of competing risks from concomitant cardiovascular diseases in patients with ARAS was emphasized in an observational analysis of 68 patients with ARAS followed up for 39 months without initial revascularization. End-stage renal disease developed in 6 (8.8%) patients, but apparently was unrelated to progression of ARAS in 5. During follow-up, 19 (28%) patients died from unrelated causes, including stroke and myocardial infarction. ⁹ In the group as a whole, reduction in renal function (mean serum creatinine 1.4 mg/dL at baseline versus 2.0 mg/dL at follow-up) was modest, and systolic blood pressure remained stable during follow-up (157 mm Hg versus 155 mm Hg), albeit, with a significant increase in antihypertensive medications (1.6 versus 1.9). Four (5.8%) patients underwent renal revascularization for refractory hypertension (n = 1), progressive ARAS (n = 1), and during aortic reconstruction (n = 2). One patient (1.5%) underwent nephrectomy for blood pressure control. Deterioration of renal function and mortality risk were higher in patients with bilateral or solitary kidney ARAS.

Treatment

The primary goals of treatment of RAS are prevention of renal insufficiency and treatment of RVH. Atherosclerotic renal artery stenosis frequently coexists with atherosclerotic disease in other vascular beds. Therefore, therapy aimed at preventing atherosclerotic complications, such as aspirin, smoking cessation, and lipid-lowering medications should be encouraged strongly. Whether these therapies have any clinically significant effect on progression of ARAS is not known. Renal revascularization mechanically corrects renal stenosis and may be performed by either surgical or percutaneous techniques.

Surgical revascularization

Surgical revascularization entails either aortorenal bypass using saphenous vein grafts or transaortic endarterectomy. In a recently published series of 222 patients who underwent surgical revascularization for uncontrolled hypertension and/or preservation of renal function, long-term cure or improvement of blood pressure was seen in 74% and improvement or stabilization of renal function in 71%. However, surgery was associated with significant operative mortality (2.2%) and morbidity (postoperative thrombosis or stenosis of the repaired artery in 7.3%). ¹⁰

Percutaneous revascularization

Traditionally, percutaneous transluminal angioplasty (PTA) and percutaneous transluminal angioplasty stenting (PTAS) were considered alternatives to surgical repair. However, the significant morbidity associated with aortic surgery and anesthesia, the longer hospital stay and recovery period after surgery, and the tremendous refinements in percutaneous techniques have caused PTA/PTAS to rapidly become the preferred forms of revascularization for ARAS. For PTAS, primary success rates from 88% to 99% ^11,12 and lower restenosis rates than PTA (14% versus 48%) alone have been reported. ¹¹

Percutaneous revascularization for blood pressure control

Uncontrolled retrospective studies have shown improvement in blood pressure in 64% to 100% of patients after PTA/PTAS; however, hypertension is rarely cured after renal PTA/PTAS. ¹³ Few prospective studies have compared renal revascularization with medical therapy for the management of hypertension in patients with ARAS. In the Essai Multicentrique Medicaments vs. Angioplastie study, ¹⁴ 49 patients with ARAS were assigned randomly to PTA (n = 23) or medical therapy (n = 26). Mean ambulatory blood pressure at 6 months was similar among the groups, but was achieved with less antihypertensive medication in the PTA group than in the medical therapy group, confirming the drug-sparing effect of PTA. In 55 patients with unilateral or bilateral ARAS and uncontrolled hypertension (diastolic blood pressure >95 mm Hg despite therapy with >= 2 antihypertensive medications) randomized to PTA or medical therapy, the Scottish and Newcastle Renal Artery Stenosis Collaborative Group found a modest benefit of PTA on systolic blood pressure, particularly in patients with bilateral ARAS. ¹⁵

In the Dutch Renal Artery Stenosis Intervention Cooperative Study Group, 106 patients with hypertension and ARAS were randomly assigned to PTA (n = 56) or drug therapy (n = 50). Patients had >50% RAS with serum creatinine <2.3 mg/dL, and either a diastolic blood pressure >95 mm Hg despite treatment with two antihypertensive drugs or an increase of at least 0.2 mg/dL in serum creatinine during treatment with an angiotensin-converting enzyme inhibitor. At 3 months, blood pressures were similar in the 2 groups, although the PTA group required less antihypertensive medication than the medical therapy group (2.1 versus 3.2, P <0.001). However, 22 (44%) patients in the drug therapy group crossed over to the PTA group because of persistent refractory hypertension or deterioration of renal function. At 12 months, no significant differences were seen between the groups with respect to systolic or diastolic blood pressures, daily drug doses, or renal function. In 3 (5.4%) patients in the medical group, ARAS progressed to complete occlusion. The investigators concluded that PTA has little advantage over drug therapy in the management of hypertension in patients with ARAS. ¹⁶ The major criticism of this study has been the high crossover rate from medical therapy to PTA, which systematically underestimates benefit from renal angioplasty. Further, it has been debated that the use of PTA (rather than PTAS) represents suboptimal revascularization strategy.

Percutaneous revascularization for renal preservation

Watson and colleagues ¹⁷ studied renal function (measured as the reciprocal of serum creatinine concentration) before and after PTAS in 33 patients with renal insufficiency (serum creatinine >1.5 mg/dL) with either bilateral ARAS or solitary kidney ARAS. All patients had progressive decline in renal function before PTAS. Technical success was achieved in all 61 vessels. Among 25 patients with complete follow-up, PTAS resulted in stabilization of renal function in 18. Renal function continued to decline in the other 7 patients, albeit at a slower rate than before PTAS. Thus, the investigators concluded that in patients with bilateral or solitary kidney ARAS and renal insufficiency, PTAS stabilizes renal function.

Indications for renal revascularization

It is reasonable to recommend revascularization for patients with the following: bilateral or solitary kidney ARAS with progressive renal insufficiency; and ARAS with refractory or complicated hypertension, refractory heart failure, or unstable angina. It is debatable whether patients with severe ARAS who do not meet the above criteria should undergo PTAS.

Future directions

A clinical need exists to clarify the role of screening for ARAS during coronary angiography, and to identify clinically silent patients with ARAS who are likely to benefit from PTAS. For example, it is unknown whether PTAS should be considered in patients with unilateral ARAS with stable mild-to-moderate renal insufficiency or those with severe ARAS incidentally diagnosed on abdominal aortography performed for an unrelated indication, eg, claudication. Proponents of pre-emptive percutaneous revascularization argue that because ARAS treated medically is frequently progressive (often in the face of adequate blood pressure control) and can cause total occlusion in a significant minority, it should be sought out aggressively and revascularized to avoid renal insufficiency. Opponents support cautious restraint, given the lack of adequate outcome data to support this strategy; the small, but real, risk of complications from PTAS (such as atheroembolization, worsening renal function, access problems); and the restenosis rate of 10% to 20%.

Patients with ARAS who have concomitant parenchymal disease, manifesting as high renal resistance index on renal artery Doppler ultrasonography, increased serum creatinine, or proteinuria, do not fare as well after PTAS as do patients without concomitant parenchymal disease. ¹⁸ However, it is not known whether PTAS or medical therapy alone is the better treatment strategy in such patients.

Whether lateralization indices (ie, measures of split renal function) and the hemodynamic response of the renal vasculature to intrarenal vasodilators (such as fenoldopam) will aid decision making in patients with ARAS is the subject of ongoing studies. Among patients undergoing PTAS, the role of distal protection to prevent atheroembolic complications is also an area of intense investigation.

Percutaneous repair of abdominal aortic aneurysm

Early animal work on endovascular repair of abdominal aortic aneurysm (AAA) was conducted on experimentally created aneurysms in the mid-1980s. The first recorded human endovascular AAA repair was performed in 1990 by Juan Parodi, ¹⁹ using a customized tube graft made of Dacron (DuPont Pharmaceuticals Corp.; Wilmington, DE) and a Palmaz-type stent at the proximal end. Lack of distal fixation was recognized to be a significant problem; hence, a distal stent graft was added with improvement in success rates. Because most AAAs required treatment of the aortic bifurcation as well, the graft design gradually evolved into the current generation of modular bifurcation devices. Several proprietary endoluminal stent graft designs are currently available or under development (Table 4, Figure 1).

The need for endoluminal graft exclusion (ELG) implantation is driven by the relatively high incidence of AAA, the risk from ruptured AAA, and the morbidity and mortality associated with surgery for AAA. The annual risk of AAA rupture, based on AAA size, is shown in Table 5. ²¹ For AAA rupture occurring out-of-hospital, the mortality is as high as 90%. Even in patients who have good surgical risks, the operative mortality for elective AAA repair ranges from 1% to 5%.

Patient selection

Patient factors

Currently, ELG is considered for patients with significant comorbidities who are not candidates for conventional surgical repair. In high-risk candidates, the operative mortality exceeds 10%, and major complications occur in 15% to 30%. Thus, in its current role, ELG is a treatment option for poor surgical risk candidates and is not considered a replacement for conventional repair. ²² However, the enthusiasm for ELG repair in such patients is tempered by the fact that these patients have high mortality secondary to their comorbidities; in one study, the 1-year survival of patients considered unfit for conventional surgery was only 20% after ELG implantation. ²³

Technical factors

Several anatomic considerations affect the suitability for ELG implantation and selection of the appropriate device. These include the proximal neck diameter and length, the length from the renal arteries to the aortic bifurcation, the distal neck diameter and length, involvement of the iliac arteries, iliac artery attachment diameter, and iliac and femoral diameters. Of particular note, tortuosity of the aorta is associated with increased complexity of ELG implantation as well as greater risk of endoluminal leaks. Computed tomographic angiography with three-dimensional reformatting and abdominal aortography are the tests most frequently used to delineate aortic anatomy before ELG implantation.

ELG designs

Briefly, three different types of aortic stent grafts have been used, depending on the inferior extent of the aneurysm: bifurcation grafts, aortoiliac grafts, and straight tube grafts (Figure 1). In a large multicenter study of 899 patients undergoing elective ELG repair, bifurcated devices were used in 91% of patients, straight tube grafts in 7%, and aortoiliac grafts in 2%. ²²

Clinical results

The AneuRx Stent Graft System (Medtronic, Inc.; Minneapolis, MN) and Ancure Endograft System (Guidant Corp.; Santa Clara, CA) were approved for clinical use in the United States. However, due to many device malfunctions and adverse events, Guidant halted production and announced a recall of all existing inventory on March 16, 2001. The combined experience from AneuRx phase I to phase III trials includes 1192 patients enrolled from June 1996 to November 1999. Technically successful AAA repair was performed in 98% of patients. At 3-year follow-up, the incidence of rupture was 0.8%. Most cases of rupture occurred in the 174 patients treated with an early stiff-graft design used in phase I and the early part of phase II. Among the patients who underwent ELG implantation using the commercially available segmented bifurcation stent graft, freedom from aneurysm rupture was 99.5% at 3 years. At 1, 2, and 3 years after implantation, cumulative patient survival was 93%, 88%, and 86%, respectively; freedom from conversion to open repair was 98%, 97%, and 93%, respectively; and freedom from secondary procedures was 94%, 92%, and 88%, respectively. The presence or absence of endoleak on contrast CT scanning was not a predictor of long-term outcome. ²⁵ Among 230 patients enrolled in the AneuRx phase II trial, the incidence of blood loss, transfusion requirements, time to extubation, number of days spent in the intensive care unit, length of hospital stay, and recovery period until return to normal function were all significantly lower in the percutaneous group compared with the surgical group.

In the Ancure phase II trials, ²⁶ 542 patients who were acceptable surgical candidates were enrolled to receive 153 tube, 268 bifurcated, and 121 aortoiliac stent grafts. Conventional surgery was performed in 111 patients who served as controls. The device deployment success rate exceeded 90% for all stent configurations. Compared with patients treated surgically, those who received the tube or bifurcated stents had similar operative mortality and bowel and wound complications, but had lower respiratory, cardiac, and bleeding complications. Patients who received the bifurcated stents had a higher incidence of renal insufficiency compared with surgical controls. Data on complications in the patients who received the aortoiliac stent design were not reported. At 12-month follow-up, shrinkage in AAA size (defined as >= 5 mm reduction in the largest AAA diameter on computed tomography scan) was seen in more than half the patients with ELG implantation (43% tube, 51% bifurcated, 58% aortoiliac stents), whereas increase in AAA size occurred in 2% to 6%. Late surgical conversion and proximal stent migration were extremely rare, and none of the patients suffered AAA rupture.

Potential complications of endoluminal AAA repair

The potential complications of AAA repair are listed in Table 6. Perhaps the most significant complication is endoleak, defined as persistence of blood flow outside the graft but within the lumen of the aortic aneurysm sac. Four types of endoleaks have been defined according to the site of leak (Figure 2). In a meta-analysis of 1189 patients enrolled in 23 studies, transfemoral ELG implantation was performed successfully in 1118 patients. ²⁷ Endoleaks were present in 270 (24%) patients at the time of hospital discharge, and involved the distal stent attachment in more than one third. Tube grafts had a higher incidence of endoleaks than did bifurcated and aorto-uni-iliac grafts. Self-expandable grafts were associated more often with endoleaks than were balloon-expandable grafts. In another study, the incidence of endoleak at 18-month follow-up was 10%. ²⁴

Endoleaks increase the pressure inside the aneurysm sac, cause an increase in the aneurysm diameter, and may predispose patients to aneurysm rupture. Periodic surveillance of ELG implants with CT scanning is mandatory to diagnose persistent endoleaks. Types I and III endoleaks are generally treated with insertion of a new ELG implant or surgical conversion. Type II endoleaks, related to retrograde flow through patent lumbar or mesenteric branches, may be treated by the selective injection of coils or a thrombogenic copolymer, eg, ethylene-vinyl-alcohol copolymer.

A recent report highlighted the increased incidence of midterm and long-term complications after elective ELG implantation in 239 patients followed up at 2 centers for 16 months: death (from unrelated causes) 22%, AAA rupture 0.8%, type I endoleak 2.9%, type II endoleak 5.4%, type III endoleak 0.4%, aortoduodenal fistula 0.8%, graft thrombosis or stenosis 2.9%, and secondary intervention or surgery 10%. ²⁸ It should be recognized that this report represents several devices used since 1992 by multiple operators in high-risk patients, and may not represent the current state-of-the-art technology.

Future issues

Widespread acceptance of ELG implantation and the preferential use of ELG over conventional surgical repair in patients who are candidates for both forms of therapy will require meticulous assessment of long-term clinical outcomes, quality of life, and cost. Of particular note is the modification in device designs to reduce the incidence of endoleaks, given the excess morbidity and mortality rates associated with their surgical treatment. Endoluminal graft exclusion implants constructed with thrombus-resistant material or biomembrane coatings (such as phosphorylcholine) or those coated with pharmacologic agents (such as antiproliferative agents to reduce intimal hyperplasia or heparin to prevent thrombus formation) that elute into the vessel wall are currently under development. ²⁹ Similar technology has been used successfully in coronary stents in patients with coronary artery disease, with improvement in angiographic and clinical outcomes.