NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Balk E, Raman G, Chung M, et al. Comparative Effectiveness of Management Strategies for Renal Artery Stenosis [Internet]. Rockville (MD): Agency for Healthcare Research and Quality (US); 2006 Oct. (Comparative Effectiveness Reviews, No. 5.)

Cover of Comparative Effectiveness of Management Strategies for Renal Artery Stenosis

Comparative Effectiveness of Management Strategies for Renal Artery Stenosis [Internet].

Show details

4Summary and Discussion

The following table summarizes the main findings that address the three Key Questions. Discussion regarding the report follows.

Table 13

Summary of Comparative Data in Treatments of Renal Artery Stenosis

Key QuestionsStrength of evidenceSummary/conclusion/comments
Key Question 1: Comparisons
Angioplasty with or without stent vs. medical treatment N/A
  • 2 RCTs evaluated long-term outcomes comparing angioplasty without stent placement to various medical treatments; 6 nonrandomized prospective or retrospective studies compared angioplasty (with or without stent) or surgical revascularization to various medical treatments.
  • 20 prospective cohorts that met criteria evaluated angioplasty with stent placement; 4 cohort studies evaluated angioplasty with or without stents.
  • Studies that compared stent placement to no stent placement found no difference in outcomes.
  • 3 cohort studies evaluated different antihypertensive medical treatments; no studies evaluated anti-hyperlipidemia or lipid-lowering drugs; 8 cohort studies evaluated the natural history of patients with RAS, on various management regimens.
  • 1 RCT, 3 nonrandomized comparative studies, and 31 cohort studies of various interventions suggest no difference in mortality up to about 5 years between revascularization and medical treatment.
 Kidney functionAcceptable
  • 2 RCTs found no difference in kidney outcomes, mostly at 6 and 12 months.
  • Among 7 other comparative studies, most found no difference in kidney outcomes, although 2 found some supporting evidence for better kidney function after angioplasty (with or without stent).
  • The cohort studies mostly support the conclusion that kidney outcomes are similar with either angioplasty or medical treatment, although improvements in kidney function were reported only among the angioplasty cohort studies.
 Blood pressureAcceptable
  • The 2 RCTs both found some evidence of greater blood pressure improvement after angioplasty than with medical treatment, although this relative effect may be limited to patients with bilateral disease.
  • Most other comparative studies found larger blood pressure reductions among patients having revascularization than medical treatment alone, although the difference was often clinically small and statistically nonsignificant. However, 2 studies found larger reductions in blood pressure among patients treated without revascularization, although the differences were not statistically significant.
  • Among cohort studies, larger reductions in blood pressure were found among medical treatment or natural history studies than in angioplasty studies, although the effect of pre-angioplasty antihypertensive medication use cannot be corrected for. Only in cohort studies of angioplasty were patients cured of hypertension, no longer requiring medication to maintain normal blood pressure.
  • 1 RCT found similar rates of cardiovascular events at 3 to 54 months of followup after angioplasty or with continued medical treatment.
  • Reporting of cardiovascular outcomes was too sparse among studies to make meaningful indirect comparisons.
 Adverse eventsN/A
  • The evidence does not support meaningful conclusions about relative adverse events or complications from angioplasty compared to medical treatment.
Key Question 2: Baseline predictors of outcomes
Angioplasty with or without stent vs. medical treatment Weak
  • In one RCT, patients with bilateral disease had larger decreases in blood pressure after angioplasty compared with medical treatment, in contrast to patients with unilateral disease.
Angioplasty N/A
  • 5 comparative studies and 15 cohort studies analyzed baseline variables as possible predictors of outcomes. Most of the comparative studies, however, did not distinguish between interventions in these analyses.
 Baseline kidney functionAcceptable
  • The 10 studies that evaluated baseline kidney function generally found that poorer kidney function (with a wide range of definitions) predicted higher mortality, poorer clinical outcomes including cardiovascular events, and/or poorer blood pressure control. However, among 4 studies, 2 found that kidney function after angioplasty improved more among patients with worse baseline kidney function, 1 found no difference in effect among patients with different baseline kidney function, and 1 found less improvement in kidney function among patients with worse baseline kidney function.
 Baseline RAS severityWeak
  • 4 studies evaluated baseline percent stenosis. The studies were heterogeneous in their analyses and their conclusions. 1 found a borderline increase in mortality among patients with >70% stenosis. 1 found that higher percent stenosis was associated with higher blood pressure after revascularization. 1 found no association with either kidney function or diastolic blood pressure. 1 found that patients with higher grade stenosis had greater benefits in their kidney function than patients with lower grade stenosis.
  • 11 studies evaluated whether bilateral vs. unilateral RAS was a predictor of outcomes. The studies were heterogeneous in their analyses and their conclusions. 2 found bilateral disease was associated with increased mortality, but 2 found no association (although 1 of these did find an association with a combined poor clinical outcome). Among 7 studies, most found no association with either change in kidney function or blood pressure, but 2 found that patients with bilateral disease had better improvement in blood pressure, and 1 found better improvement in kidney function than patients with unilateral disease.
 Baseline cardiovascular diseaseAcceptable
  • Among 6 studies, a range of cardiovascular measures, including history of disease, were found to be associated with increased risk of death, new cardiovascular events, or decreased likelihood of improvement in kidney function after revascularization. 2 studies, though, found that some baseline cardiovascular factors, including history of myocardial infarction, CHF, or hyperlipidemia, or reduced ejection fraction, did not predict increased mortality.
 Diagnostic testsWeak
  • 3 diagnostic tests were evaluated by 4 studies. The captopril test, renogram, and unilateral renin secretion were not associated with differential outcomes in blood pressure, kidney function, or mortality. 2 studies evaluated a resistance index of over 80%; 1 found that these patients had worse kidney and blood pressure outcomes and 1 found that they had better changes in both kidney function and blood pressure levels.
  • Among 5 studies evaluating age, 1 found that older patients had higher followup blood pressure, 1 that they had lower followup blood pressure, and 3 found that after adjustment for other predictors, age was not associated with poor clinical outcomes.
  • Among 3 studies evaluating sex, 2 found that men had worse outcomes than women, but 1 found no difference after adjustment for other predictors.
Medical treatment N/A
  • No study evaluated potential predictors of outcomes.
Natural history N/A
  • 4 natural history studies examined various predictors, 2 of which performed multivariate analyses.
 Baseline kidney functionWeak
  • 1 study found that lower baseline GFR was independently associated with higher mortality or dialysis.
 Baseline RAS severityWeak
  • 2 studies found that higher grade stenosis was independently associated with higher mortality (1 by multivariate, 1 univariate analysis); 1 study found that bilateral disease was not associated with kidney disease prognosis.
 Baseline cardiovascular diseaseWeak
  • 1 study found that various markers of cardiac disease predicted mortality in patients with coronary artery disease and RAS.
 Diagnostic testsWeak
  • 1 study found that patients with nonspiral blood flow in the renal arteries had significant progression in kidney impairment, while those with spiral flow did not.
  • 1 study found that older age predicted mortality in patients with coronary artery disease and RAS.
Key Question 3: Effect of periprocedural interventions on outcomes
Angioplasty with or without stent Weak
  • 2 studies found no difference in blood pressure and kidney outcomes between patients who had stents placed and those who did not.
Other interventions N/A
  • No study that met eligibility criteria reported analyses of whether other periprocedural interventions, such as different drugs or different approaches, affected either complications or long-term outcomes.

Abbreviations: CHF = congestive heart failure; GFR = glomerular filtration rate (or creatinine clearance); N/A = not applicable; RAS = renal artery stenosis; RCT = randomized controlled trial.

As evidenced from discussion among nephrologists, surgeons, interventional cardiologists and radiologists, and other experts, in addition to perusal of both review articles and primary studies on management of atherosclerotic renal artery stenosis (ARAS), there remains uncertainty about the best specific interventions for patients; although the American College of Cardiology and the American Heart Association have issued clinical guidelines on management of renal artery stenosis (RAS). These guidelines are based in part on evidence also included in this review, in addition to retrospective and small studies that did not meet this review’s eligibility criteria, and expert opinion.

A number of issues complicate the process of making decisions both for individual patients and for populations of patients. For one, the exact definition of ARAS varies depending on which diagnostic test is used, what threshold for stenosis is preferred, what degree of either resistant hypertension or of kidney damage is required, and whether other evidence of atherosclerotic disease is present. Furthermore, the definition and relative importance of these items have been and continue to change as new diagnostic tests are used or existing tests are refined, as definitions of chronic kidney disease change, as treatments for hypertension improve, and also as techniques and modalities of surgical and percutaneous interventions change and, presumably, improve. In addition, for individual patients, the evaluation of RAS may be complicated by the risks, difficulties, and expense of the diagnostic tests. Each diagnostic test has potential limitations related to operator skill, their invasive nature, risks due to contrast dye, or lack of availability, in addition to the use of various thresholds for and definitions of RAS.

The challenge of treating ARAS to achieve the targeted outcomes of improved blood pressure control and preservation of kidney function lies in the significant overlap between etiologic factors of aortorenal vascular disease and parenchymal kidney disease. While diabetes mellitus, dyslipidemia, and elevated blood pressure are associated with atherosclerotic narrowing of the renal arteries and consequent worsening of blood pressure and kidney function, they are independently associated with direct kidney injury. In a great many cases, overcoming the renal artery lesion fails to improve hypertension or kidney function, which may be mediated not only by ARAS but also by underlying kidney disease. Systematically evaluating the role of ARAS in hypertension and kidney dysfunction will assist in determining whether intervention should be directed towards improving kidney perfusion through angioplasty with stent placement or more aggressively targeting the underlying factors of parenchymal kidney disease with combination medical therapy.

For individual patients and their clinicians the question of what the preferred treatment for ARAS may be is fraught with difficulties largely related to the frequent frailty of these patients and the known complications from any of the interventions. These patients are generally elderly, often with severe cardiovascular disease including atherosclerosis and diastolic left ventricular dysfunction, often with moderate or severe chronic kidney disease, and with diabetes. Each of the antihypertensive agents carries substantial risks of bothersome and dangerous adverse events, which may be more likely or serious when multiple drugs are used. These drugs in general need to be taken lifelong and may only prevent further worsening of cardiovascular or kidney disease, as opposed to lessening the severity of disease. Invasive interventions, whether open or percutaneous, however, also carries risks of immediate death, cardiovascular events, kidney damage, and pain, or other effects on quality of life. Also, the procedure may not carry any noticeable benefit to patients, in that they are likely to continue to require antihypertensive medications and may have no survival, cardiovascular, or kidney benefit. Thus the relative overall effectiveness of angioplasty and continued aggressive medical treatment for most patients with ARAS remains unclear. For some patients with acutely worsening kidney or cardiovascular function, anecdotal evidence strongly suggests a benefit to revascularization; however, very few studies explicitly include such patients. Thus this review is not applicable to patients with clinical conditions necessitating acute intervention.

In 1993, the 5th Joint National Committee on Detection, Evaluation, and Treatment of High Blood Pressure (JNC-5) came out with recommendations that placed greater emphasis on attempting to achieve lower blood pressure levels than earlier sets of recommendations had made. This coincided with the increased use of angiotensin converting enzyme (ACE) inhibitors, and subsequently angiotensin receptor blockers (ARBs), which for many patients were both more effective and better tolerated than other drugs for reducing blood pressure, particularly when used in combination with some of the other drugs. Thus, in the early to mid-1990s many patients with previously “resistant” hypertension could now be better controlled, whether they had RAS or another cause of hypertension.

At about the same time, percutaneous angioplasty began to be more commonly used to revascularize patients’ stenotic renal arteries instead of major open surgical techniques. Also as stent placement has become more common for atherosclerotic coronary and other arteries, stents also have been more commonly placed during renal artery angioplasty. This shift can be seen in the literature, where the majority of cohort studies on angioplasty did not use stents (or at least did not report using stents), while 80 percent of the cohort studies that included patients treated since 1993 did employ stents.

These changes, however, have been occurring in an era when there has been little high quality evidence (prospective comparative trials) to support the relative benefit of angioplasty, with or without stents, compared to aggressive medical treatment. While the theoretical benefits of revascularization are appealing, there is no robust evidence to allow individual patients and clinicians to decide which treatment option is best.

For this reason, the CORAL trial has been designed to address both whether clinical benefits are greater with angioplasty with stent placement or aggressive medical treatment, and to determine which patients may benefit most from one intervention or the other. However, currently the evidence base includes two relatively short duration randomized trials of moderate methodological quality that compared angioplasty, mostly without stent placement, to a wide variety of antihypertensive treatment.

The two trials evaluated only 103 patients, who at baseline had ARAS of greater than 50 or 60 percent, only 16 of whom had bilateral disease, and about half of whom had ostial disease. Their blood pressure prior to the studies was generally poorly controlled with mean blood pressures ranging from 165–190/96–105 mm Hg. Even after treatment, on average their blood pressures remained elevated at approximately 151–187/88–103 mm Hg. It is difficult to assess from the reports, but it appears that only a small minority of patients were treated with ACE inhibitors or ARBs. In one study, the mean serum creatinine was under 2.0 mg/dL, probably implying stage 2 or 3 chronic kidney disease. In the study restricted to patients with unilateral disease, patients may have had better kidney function, with a mean creatinine clearance of 73 mL/min (stage 2 chronic kidney disease).

The two trials found no difference in kidney function or progression to end stage renal disease, or (in one study) cardiovascular event rates. The effects on blood pressure are mixed. One study found a substantially greater benefit on blood pressure in those patients with bilateral disease who had angioplasty compared to those who did not (−34/−11 vs. −8/−1 mm Hg), but no difference among patients with unilateral disease. In the other trial of only patients with unilateral disease, both diastolic and systolic blood pressure decreased by 7 mm Hg more after angioplasty than with medical treatment, but only the change in diastolic pressure was statistically significant. However, after angioplasty, patients took only half as many antihypertensive drugs as those who continued on medical treatment. Though, on average, patients in both arms remained hypertensive (151/90 and 158/95 mm Hg).

The CORAL study in contrast is enrolling patients with over 60 percent stenosis, poorly controlled hypertension on two or more drugs, but not chronic kidney disease. It will also be comparing interventions that are more current than the two trials published in 1998, including angioplasty with stent placement, the antiplatelet agent clopidogrel, and the ARB candesartan. The two published randomized controlled trials (RCTs) that compare angioplasty to medical treatment alone used somewhat different eligibility criteria that imply inclusion of patients with different severity of ARAS compared with patients being enrolled in CORAL. One RCT used similar criteria for percent stenosis, but only in patients with unilateral disease; blood pressure and kidney function criteria were narrower, suggesting that on average hypertension and kidney disease were less severe. The other RCT included patients with lower grade stenosis (>50 percent), but did not exclude patients with more severe hypertension and included patients with more severe kidney disease. Among the remaining studies that compared revascularization to medical treatment and the noncomparative cohort studies, there were a wide range of eligibility criteria, commonly including patients with stenosis as low as 50 percent, or with either more or less severe blood pressure and kidney function. Across studies there was no clear evidence that differences in eligibility criteria were predictive of outcomes – except possibly that patients with bilateral disease had greater improvement after angioplasty, compared to those with unilateral disease. However, it was evident, by comparing mortality rates or change in kidney function across studies, that studies did differ in the severity of disease among their enrolled patients; although, eligibility criteria such as percent stenosis, blood pressure, kidney function, and others were not clearly associated with overall outcomes. Furthermore, the evidence does not adequately address how differences in eligibility criteria may affect the comparison between angioplasty and medical treatment.

The remainder of the current literature consists of randomized trials comparing immediate to delayed or no revascularization, or comparing surgical revascularization to medical treatment, prospective and retrospective nonrandomized comparative studies, and prospective and retrospective uncontrolled cohort studies. Gleaning comparative effectiveness from these studies is fraught with numerous biases due to lack of randomization (among the large majority of these studies) and poor applicability. It is highly likely in many of these studies that patients were chosen either for revascularization or for medical treatment based on many factors separate from their ARAS alone including age, comorbidities, severity of symptoms or of associated conditions, clinician preferences, and others.

Assessing the applicability of these studies to the population being enrolled for the CORAL study is also problematic, both because of the same biases discussed and because, as discussed above, the definition of ARAS, the diagnostic tools used, and the interventions employed have changed both subtly and greatly over the past 15 years that make up the bulk of this review. One place where the literature review theoretically can be helpful to the current stage of the CORAL study is in estimating the power needed to address the primary and secondary outcomes and planned analyses. However, this review has found great heterogeneity in all outcomes assessed across studies, with little or no indication what the specific causes of the heterogeneity are. As an example the mortality rates across studies vary from nil to 80 percent at various time points over the first 5 years of followup. It is probably a truism that those studies with higher mortality rates included sicker patients (or possibly more poorly treated patients), reviewing the available data it is unclear which factors at baseline would have predicted mortality rates in any given study.

Another limiting issue was that adverse event reporting was generally sparse and not reported in a consistent manner. Revascularization studies tended to focus exclusively on periprocedure complications, without considering any RAS-related drug adverse events. Natural history studies did not report any adverse events. Even the adverse events reported by drug studies were incompletely reported. In particular, none of the studies addressed complications or adverse events in a manner that could allow comparison of risks between the two interventions, except one study that reported 30-day mortality.

Regarding Key Question 2, on the value of baseline factors for predicting clinical outcomes after either revascularization or continued medical treatment, few studies performed adequate multivariable analyses, controlling for the many confounding factors. In addition only one comparative study attempted to determine which baseline variables might predict a better outcome with one intervention or the other. This study concluded that the benefit of angioplasty over medical treatment in reducing blood pressure was confined to those patients with bilateral disease. Also, very few studies evaluated the value of diagnostic tests to predict outcomes. None analyzed whether any diagnostic tests would predict a better outcome with alternate treatments, except for the RCT comparing immediate versus delayed or no revascularization, where the captopril test and renogram did not predict outcomes.

The question of whether any procedure-related variables might affect complication rates or long-term outcomes was addressed by only a few studies that compared stent placement to no stent placement, where no difference was found. Among the studies that met eligibility criteria, no study evaluated any procedure-related drug or technique. In addition, no study evaluated any drugs other than antihypertensive agents, such as antilipid or antiplatelet drugs.

In conclusion, there is no published evidence directly comparing angioplasty with stent placement and “aggressive” medical treatment with currently available drugs for ARAS. Overall, the evidence does not currently support one treatment approach over the other for the general population of people with ARAS. Notably, almost two-thirds of the studies were of poor methodological quality and more than half were of limited applicability to the population of interest. A very limited evidence base directly compares angioplasty without stent placement and medical treatment. While there was a benefit in blood pressure measurements after angioplasty, particularly in patients with bilateral disease, there was no difference in kidney function outcomes, and possibly no differences in mortality and cardiovascular event rates, although studies generally were included too few patients and were of too short a duration to make definitive assessments regarding these clinical event outcomes. Comparison of adverse events and complications across the various interventions is difficult. However, it is clear that various complications after revascularization do occur in a small percentage of patients, and each of the antihypertensive drugs has associated adverse events. Among the studies reviewed, the predictive value of diagnostic tests either for long-term outcomes or to help determine the best treatment is unclear. A variety of indicators of the severity of ARAS or of health problems, such as poorer kidney function, worse blood pressure, and coexisting cardiovascular disease predict poorer outcomes in patients with ARAS. The reviewed studies did not report any indicators that may predict improved outcomes. Very limited evidence from direct comparisons suggests there is no difference in outcomes based on whether patients had stents placed or not. The studies that met eligibility criteria (generally larger and/or prospective studies, excluding case reports and series) did not address the effect of any other procedure-related intervention. As the reviewed studies did not explicitly address the population of patients who may need acute intervention because of rapid clinical deterioration, the conclusions of this review do not apply to these patients.

PubReader format: click here to try


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...