Atherosclerotic Renal Artery Stenosis: To Treat Conservatively, to Dilate, to Stent, or to Operate?

Prevalence

The true prevalence of ARAS in unselected patients is unknown. Early autopsy studies reported incidences of severe ARAS (lumen diameter reduced by >50%) of about 25% in subjects aged 50 yr or over (10). More recent prevalence estimates differ greatly according to clinical presentation. In the hypertensive population, the prevalence of diagnosed ARAS for which revascularization is considered (i.e. lumen diameter reduced by >50% (2,6,8,9) or 60% (7)) is probably <5%. Critical (≥60%) stenoses were reported in 10 (22%) of 45 patients undergoing renal duplex ultrasound at the start of renal replacement therapy, 7 (70%) of whom had associated cardiovascular disease (10). ARAS typically occurs in high-risk patients with coexistent vascular disease elsewhere (2,5,11,12,13,14,15). Most patients are men, past or present smokers, one in two has high serum cholesterol levels, and one in five has diabetes mellitus (11). ARAS is frequent in patients with coronary heart disease (11,12) or a history of stroke (13). It is particularly frequent in elderly patients with peripheral vascular disease (14) or congestive heart failure (15) and may be found in the absence of hypertension (Table 1). ARAS is frequently bilateral and associated with renal failure in patients with acute congestive heart failure (5,15).

Renal Survival of Patients with ARAS

Atherosclerotic renovascular disease is an increasingly important cause of end stage renal disease (ESRD) (2,16,17). Although the precise risk and frequency are unclear, ESRD is usually a consequence of bilateral ARAS and/or of associated parenchymal disease. ARAS is bilateral in almost half the patients with renovascular disease (2) and may affect a solitary kidney in cases with occlusion on the contralateral side. Actuarial renal survival differs greatly between patients with unilateral stenosis, those with bilateral stenosis, and those with unilateral stenosis plus contralateral occlusion. In a prospective cohort study, two-year renal survival of patients in these three categories was found to be 97%, 82%, and 45%, respectively (18). Parenchymal disease associated with ARAS includes consequences of aging, hypertension-induced nephrosclerosis, diabetes (when present), and renal injury related to atheroembolism (16,17). It is usually associated with proteinuria (19,20).

The progression of renal failure in patients with ARAS may reflect progression in the degree of narrowing of the renal artery, progression in associated parenchymal disease, or both. Progression in renal artery stenosis has been documented in retrospective angiographic studies and, more recently, in a large prospective study using renal artery duplex ultrasound. Five reports concerning serial angiograms in 237 patients with ARAS were reviewed by Rimmer and Gennari (2). Progression, including worsening of existing stenosis of the renal artery or the development of contralateral renal artery stenosis, was reported in 116 patients (49%) during follow-up periods of 6 to 180 mo. Renal artery occlusion occurred in 28 cases (14%). These figures for progression and occlusion were probably overestimated by a selection bias as patients in these studies often had repeated angiography because of signs or symptoms of worsening vascular disease. In a recent cohort study, 170 patients (295 kidneys) with at least one ARAS were monitored with serial renal artery duplex scans for a mean of 33 mo (21). The 3-yr cumulative incidence of progression, defined as any detectable increase in the degree of diameter reduction affecting at least one renal artery, was 35%. In renal arteries initially classified as normal, <60% stenosis, and ≥60% stenosis, the 3-yr incidence rates were 18%, 28%, and 49%, respectively. There were only 9 (3.1%) renal artery occlusions, all of which occurred in arteries with 60% stenosis at the examination before the detection of occlusion. In a stepwise Cox proportional hazards analysis, baseline risk factors associated with progression were a systolic BP ≥160 mm Hg, diabetes mellitus, and high-grade (≥60% stenosis or occlusion) ipsilateral or contralateral stenosis (21).

It is unknown whether progression in parenchymal disease is correlated with progression in renal artery stenosis. In another report from the cohort discussed above, renal atrophy was defined as a reduction in renal length at duplex scan of >1 cm (22). The 2-yr cumulative incidence of renal atrophy was 5.5%, 11.7%, and 20.8%, respectively, in renal arteries initially classified as normal, <60% stenosis, and ≥60% stenosis. The occurrence of atrophy was correlated with changes in serum creatinine concentration, suggesting a relationship between renal artery disease and parenchymal disease. Parenchymal disease may progress to ESRD, however, despite successful revascularization (16,17,23).

Survival of Patients with ARAS

Cardiovascular events are much more frequent than ESRD in patients with ARAS. In a Swedish survey of 164 consecutive patients with renal artery stenosis of ≥50%, the risk ratio for overall mortality was 3.3 when using the normal population of Sweden, matched for age, as a reference, whereas the risk ratio for cardiovascular mortality was 5.7 (24). Of the 44 patients who died during the 7.1-yr follow-up period, 33 died from cardiovascular diseases, 9 died with noncardiovascular nonrenal diseases, and only 2 progressed to ESRD. The overall risk ratio for mortality in patients with renal artery stenosis was higher than that of patients hospitalized for stable angina and similar to that of patients undergoing surgery for colon cancer.

Medication

Long-term therapy using diuretics, beta blockers, angiotensin-converting enzyme inhibitors (ACEI), aspirin, and statins reduces mortality from coronary heart disease, congestive heart disease, and stroke in high-risk hypertensive patients, particularly in elderly patients and in those with diabetes mellitus or a previous cardiovascular event. As mentioned earlier, most patients with ARAS have hypertension and associated cardiovascular risk factors, and many have symptomatic atherosclerosis elsewhere. They should be provided with pharmacologic treatment according to current recommendations, and those who smoke should be advised to quit.

Special attention should be paid to ACEI, however, because they can induce renal dysfunction in patients with bilateral stenosis, stenosis in a solitary kidney, or severe nephrosclerosis (16,17,18). Short-term treatment with ACEI is safe in patients with ARAS, provided plasma creatinine concentration is closely monitored with ACEI treatment stopped if plasma creatinine increases by 20% or more. In 108 patients at high risk for severe ARAS, 44 of whom had bilateral stenosis, 29 had a solitary functioning kidney and 20 had unilateral stenosis (25). ACEI without diuretics caused at least a 20% increase in plasma creatinine concentration in 26 cases within 4 d and in another 31 cases after 4 wk. The increase in plasma creatinine concentration during ACEI was correlated with the severity of renovascular disease. No case of acute renal failure occurred, and plasma creatinine concentration returned to normal in all cases after stopping ACEI. ACEI is probably safe in the long term in subjects with low-grade (<60%) stenosis (21,22) and in whom plasma creatinine concentration does not change during the first month of treatment (25). In addition to plasma creatinine concentration monitoring, kidney length should be determined yearly because individual kidney function may be reduced on the most stenotic side despite stable overall renal function, and this may result in progressive unilateral kidney atrophy (22). There are few reports of the renal effects of angiotensin receptor blockers in patients with ARAS, but experimental data suggest that they carry the same risk of renal dysfunction or atrophy as ACEI.

Revascularization Procedures and Procedural Outcome

PTRA, with or without stenting, and surgical reconstruction are the two options for renal revascularization. PTRA is currently the first choice option because a randomized trial in patients with ostial ARAS has shown that it is simpler than and as effective as surgical reconstruction (3). Surgical bypass is currently reserved for patients in whom PTRA and stenting fail and in hypertensive patients with ARAS who require infrarenal aortoiliac reconstruction. Prophylactic renal artery repair is not justified, however, in patients with ARAS and aortoiliac disease but without hypertension or renal failure (26).

Problems associated with PTRA alone include elastic recoil causing immediate failure and late restenosis, both of which are more frequent in cases of ostial stenosis than in nonostial stenosis (4,27). Stent placement after PTRA prevents immediate recoil but does not completely eliminate restenosis. Leertouwer et al. (4) reviewed 14 uncontrolled studies of renal artery stenting (Table 2). Most stented arteries had ostial stenoses. The overall incidence of restenosis in patients with angiographic follow-up was 17.0%. van de Ven et al. (8) compared PTRA plus stenting and PTRA alone in 84 patients with ostial ARAS. The primary outcome measure was the success rate of the primary procedure. The immediate success rate was higher, the 6-mo restenosis rate was lower, and the 6-mo patency rate was consequently higher (80% versus 51%) after PTRA plus stenting than after PTRA alone. The nature and rate of complications did not differ between the two procedures.

BP Outcome After Revascularization

Three randomized trials in patients with ARAS have compared PTRA with antihypertensive medication to medication alone. Webster et al. (6) randomly assigned 55 patients with unilateral or bilateral ARAS to two groups: intervention or medication alone. Primary outcome measures were the mean changes from baseline in office BP and in serum creatinine concentration at 6 mo. In unilateral cases, no significant difference in BP was observed between the intervention and medication groups, In bilateral cases, 6-mo BP changes were similar in patients undergoing intervention and those given medication alone. The changes in BP at the most recent follow-up (3 to 54 mo) were compared. A statistically significant difference in systolic pressure was observed (-34 and -8 mmHg in the intervention and medication groups, respectively). Plouin et al. (7) randomly assigned 49 patients with unilateral ARAS to two groups: PTRA (with stenting if deemed necessary) or medication alone. The primary outcome measure was the mean change from baseline in 24-h ambulatory BP at 6 mo. The average reduction in BP was similar for the two groups, but PTRA reduced by 60% the probability of having a treatment score of ≥2 at termination (P < 0.001). In the trial carried out by van Jaarsveld et al. (9), 106 patients with ARAS were randomly assigned to two groups: PTRA (with stenting if deemed necessary) or medication alone. The primary outcome measures were systolic and diastolic office BP at 3 and 12 mo after randomization. The mean reductions in BP were similar in the PTRA and medication groups, but the final treatment score was significantly lower in patients who had undergone PTRA. van Jaarsveld et al. (9) found that abnormal scintigraphic findings and the severity of stenosis, both assumed to be predictors of renovascular hypertension, did not predict the BP response. Overall, differences in final BP between patients treated by PTRA and by medication in these trials were minimal, and only a minority of patients undergoing PTRA were able to discontinue medication. Nevertheless, the number of antihypertensive agents required to control BP adequately was lower after PTRA than for medication alone. This is an advantage of PTRA over conservative treatment in patients with resistant hypertension.

No randomized study comparing renal artery stenting to medication has been published. In a trial comparing PTRA plus stenting to PTRA alone (8), subsequent outcome was only weakly correlated with renal artery patency; the final patency rate was higher after stenting than after PTRA alone, but the two procedures did not differ in their effects on BP or renal function. This observation is an important clinical correlate of pathologic findings. The pressure gradient across an ARAS is one component of a renovascular disease that also involves the extension of atherosclerosis to the branches of the renal artery, arteriolar thickening, tubular and glomerular atrophy, glomerulosclerosis, and the vascular consequences of antheroembolic disease and preexisting primary hypertension, which may affect both kidneys (2,16,17).

Renal Outcome of Revascularization

Randomized trials have shown no significant difference in renal function between determinations at baseline and at most recent follow-up or among patients undergoing surgery, PTRA, renal artery stenting, or receiving medication alone (Table 3) (3,6,7,8,9,28). In all treatment groups, including medication alone, the progression of renal disease as estimated by plasma creatinine concentration, was null or minimal. Follow-up was relatively short, however, and the majority of the patients had unilateral ARAS. The GFR does not reflect individual kidney function in cases with unilateral ARAS because hyperfiltration in the normal kidney may compensate for reduced filtration in the ischemic kidney. Synchronous GFR determination and kidney scintigraphy are the only means of assessing individual kidney function in cases of unilateral ARAS. Even with such techniques, PTRA does not seem to improve individual kidney function in the short term (29).

It is unclear whether renal revascularization prevents the progression of ARAS to occlusion. In the trial carried out by van Jaarsveld et al. (9), renal angiography was repeated 12 mo after randomization in 48 of the 56 patients assigned to the PTRA group and in 43 of the 50 patients assigned to the drug-therapy group. No patient in the PTRA group had total renal artery occulusion, whereas stenosis progressed to occlusion in 4 patients (9%) in the drug therapy group. No case of occlusion was reported in the trials carried out by Webster et al. (6) and Plouin et al. (7). Revascularization may prevent, but may also induce, disease progression through renal infarction and cholesterol embolism. Such complications are relatively rare but, unlike spontaneous progression, they are immediate consequences of revascularization. Uncontrolled studies assessing patients with ARAS and progressive azotemia have reported that PTRA improves renal function in 41 to 43% of patients, with a procedural mortality rate of 5 to 6% (2). A large (n = 163), long-term (up to 4 yr) study of renal outcome after renal artery stenting found that renal failure may occur despite adequate revascularization (23). Stenting was associated with 3 deaths, 21 episodes of contrast-induced renal failure, and 2 episodes of retroperitoneal hemorrhage. No nephrectomy or emergency renal bypass surgery was required, but angiographic follow-up was not reported. In a 2-yr randomized comparison of PTRA and surgery (3), one renal artery occlusion occurred in each treatment group (3.4%). In a trial comparing PTRA and stent placement (8), four patients (9.5%) in each treatment group developed clinically manifest cholesterol embolism.

There is a need for long-term comparisons of revascularization and medication to determine whether the progression rate after invasive treatment differs from that under conservative treatment. Available trials comparing revascularization and medication cannot answer this question because they lack statistical power and because they are relatively short-term. Consequently, they may overestimate procedure-induced renal complications as compared with long-term benefits. Future trials designed to assess the renal protective value of PTRA, with or without stenting, will probably involve patients with mild to moderate renal insufficiency because such patients are less prone to severe procedural complications and premature death from cardiovascular disease than those with severe renal failure (30). Conversely, the renal outcome of revascularization will be difficult to document in patients with severe renal failure.

The Decision to Revascularize

Age is a universal limitation to invasive treatment. Irrespective of the cause of secondary hypertension, the probability of BP reverting to normal after specific treatment decreases sharply with increasing age (Table 4) (31). Increasing age, as well as the presence of diabetes mellitus, is also associated with an increase in the extension and severity of atherosclerosis, which in turn increases the incidence of puncture site and renal artery complications and of cholesterol embolism. Many elderly patients with vascular disease have silent ARAS with normal or near-normal BP and renal function (Table 1). They should not be exposed to the complications of renal artery revascularization unless they develop heart failure. In hypertensive patients with ARAS and normal or near normal renal function (6,7,9), hypertension can be controlled by drugs alone in most cases and there is no clear advantage of revascularization over medication plus careful monitoring. Revascularization is clearly justified only in patients with resistant hypertension and in those who need ACEI because of a history of heart failure or myocardial infarction. Patients with ARAS and mild renal failure have a much higher risk of dying from a stroke or myocardial infarction that of progressing to ESRD (24,30). As serum creatinine level is an independent predictor of death in patients with ARAS (24,30), patients with ARAS and moderate to severe renal failure have a high risk of both ESRD and cardiovascular death. They are therefore candidates for revascularization. Unfortunately, no randomized trial has been conducted in these subjects, and retrospective reports suggest that they are at high risk of procedural complications (2). Revascularization should probably be undertaken in ARAS patients with rapidly deteriorating renal function (16) or in whom plasma creatinine concentration has increased by >20% during ACEI treatment (25).

Procedural Options

Catheter-guided angiography is a prerequisite for revascularization because it allows the precise location (unilateral or bilateral; ostial or truncal) and quantification (stenosis grade and length) of renal artery disease and provides a visual image of the aorta and of the size and architecture of both kidneys. If the kidney is <8 cm long, there is little chance of BP improvement or kidney function recovery (16). Percutaneous revascularization is the first-line method (3), surgical bypass being reserved for patients with associated aorto-iliac disease and cases with PTRA failure. Stent placement is required if there is an elastic recoil with a residual stenosis of 30% or more, which is much more frequent in ostial than in truncal stenosis. Therefore, most radiologists propose primary stent placement for stenosis within the aortic wall or within 10 mm of the aortic lumen (4,8,23). Patients are routinely given 100 to 300 mg of aspirin daily for 6 mo after PTRA, warfarin or a combination of aspirin and ticlopidine after PTRA plus stenting. Antihyper-tensive therapy is discontinued on the day of the procedure to prevent hypotension, unless it is required for cardiovascular