Renal artery stenting for atherosclerotic renal artery stenosis: A review of contemporary data

Abstract

Renal artery stenting has been less frequently performed following the results of the Cardiovascular Outcomes in Renal Atherosclerotic Lesions trial. In this trial, stenting demonstrated no benefit over medical therapy in a composite outcome including mortality as well as cardiovascular and renal events. However, these findings must be interpreted in the context of the study population, which included many patients less likely to benefit from intervention due to relatively mild renal artery stenosis, as well as less severe hypertension and renal dysfunction. More recent investigations have focused on identifying subgroups of patients who may benefit from revascularization. Identified subgroups include patients with high systolic blood pressure, diastolic hypertension, worsening kidney function, greater antihypertensive burden, flash pulmonary edema, and mild-range proteinuria. Many of these variables illustrate a clinical picture of a patient with severe ischemic kidney disease leading to problems with hypertension and volume regulation that has not yet irreparably damaged the renal parenchyma. Deciding which patients fit into this picture, and therefore should be stented, presents a challenge. This review synthesizes the current evidence surrounding these predictive variables and explores how they may be integrated into clinical decision-making.

Key Words: Renal artery stenosis; Stenting; Hypertension; Rapidly declining kidney function; Flash pulmonary edema

Core Tip: Patients likely to benefit from renal artery stenting are those with severe anatomic renovascular disease (stenosis > 80% and especially > 90% often bilaterally or in a single functioning kidney) and severe manifestations of that disease (high systolic and diastolic blood pressure despite at least 3 medications, rapidly declining kidney function with low levels of proteinuria, or flash pulmonary edema without other pathophysiologic explanations. These patients were not included in prior randomized controlled trials and represent a high-risk subset of patients with atherosclerotic renovascular stenosis. Survival benefits may be achieved especially in patients stented for rapidly declining kidney function.

INTRODUCTION

Renal artery stenting (RASt) came to prominence in the early to mid-1990’s with reports of success in treating both hypertension and renal insufficiency due to atherosclerotic renal artery stenosis[1-3]. These studies established RASt for atherosclerotic renal artery stenosis as superior to percutaneous transluminal renal angioplasty (PTRA) alone in terms of treatment response and success rate[2,4]. These studies led to a 3.7-fold increase in the number of claims filed for renal angioplasty and stenting in Medicare beneficiaries[5]. Given the increased rate of stenting, randomized controlled trials were conducted to assess the efficacy of RASt on renal insufficiency, hypertension, and a composite measure of clinical outcomes[6-8]. The Stent Placement for Atherosclerotic Renal Artery Stenosis and Impaired Renal Function[6] and Angioplasty and Stenting for Renal Artery Lesions (ASTRAL) trials[7] evaluated the effect of RASt on kidney function. The Cardiovascular Outcomes in Renal Atherosclerotic Lesions (CORAL) trial[9] evaluated the effect of RASt on a composite clinical outcome metric that involved death, stroke and cardiac and renal adverse events.

These randomized controlled trials each found no clinical benefit to RASt over optimal medical therapy but have since been subject to criticisms mainly centered on patient selection. The Stent Placement for Atherosclerotic Renal Artery Stenosis and Impaired Renal Function trial was first of the three published and found no benefit in kidney function following stenting. However, of the sixty-two patients who were randomized to the stenting group, only 46 received a stent[6]. The ASTRAL trial included 806 patients allowing for much more statistical power and again showed no improvement in kidney function following stenting compared to optimal medical management[7]. The average stenosis in the stenting cohort was seventy-six percent with thirty-nine percent of patients having fifty to seventy percent stenosis. Additionally, patients were included in ASTRAL only if there was uncertainty of whether the patient would benefit from stenting. This led to patients with severe disease being preferentially excluded. Of note, a long-term follow-up of ASTRAL data again showed no benefit of stenting over medical therapy as measured by the CORAL clinical composite outcome measure[8].

The CORAL trial attempted to include patients with more severe renovascular disease but was plagued by slow enrollment which led to an expansion of their inclusion criteria[9]. Average stenosis in the CORAL trial was 67% with a quarter of patients having their stenosis scored by duplex ultrasound. Furthermore, average systolic blood pressure (SBP) was 150 mmHg on an average of 2.1 antihypertensives with nearly 30% of patients in the stenting group already at their goal blood pressure prior to stenting[9]. Average glomerular filtration rate (GFR) in the CORAL cohort was 58 mL/minute/1.73 m² which is barely into stage 3a chronic kidney disease. The CORAL investigators addressed this by subgrouping patients based on SBP above or below 160 mmHg, GFR above or below 45 mL/minute/1.73 m², and percent stenosis above or below 80%. Even with these subgroupings there was no difference related to stenting[9]. One possible explanation for this is that each of these factors alone may not be predictive of a positive response. It should also be noted that there have been concerns raised about the validity of the stenosis measurements in CORAL given that nearly half of stented patients were reported to have stenosis > 80% despite the average stenosis being 67% with a standard deviation of 11% and minimum of 60% stenosis for inclusion in the trial[10]. Despite these limitations, the conclusion made by the CORAL investigators was that stenting does not confer benefit over optimal medical therapy.

The effect of these trials has been to drastically limit the number of endovascular procedures that are performed in patients with atherosclerotic renal artery stenosis[11]. This has likely been a positive impact for many patients who are found to have mild to moderate severity stenoses and would be less likely to benefit from stenting as was established through ASTRAL and CORAL[7,9]. However, these trials could not address whether patients with more severe renovascular disease are good candidates for RASt. The reduction in stenting has also created a challenge in adequately assessing patients with severe stenoses and medically resistant hypertension and kidney dysfunction[12]. Despite this, new studies have investigated whether patients with severe renovascular disease benefit from RASt and which predictor variables could improve patient selection. Given these remaining questions, the present review aimed to synthesize the most up to date literature on the topic of RASt and to provide an evidence-based assessment of how these variables can be used in patient selection and clinical decision making.

RAST IN THE POST-CORAL ERA

RASt has seen a steady decline in utilization following the outcomes of the CORAL trial[11,13]. The prevalence of RASt dropped by 41% from 2016 to 2020 based on Medicare claims data in patients over the age of 65[13]. This decline in usage has been matched with more strict guidelines around which patients are recommended for stenting. Patients with hemodynamically significant renal artery stenosis (either > 70% stenosis or with a significant pressure gradient across the lesion) and any of: (1) Resistant hypertension despite three anti-hypertensive medications including a diuretic; (2) Progressive ischemic nephropathy; and (3) Cardiac destabilization syndromes[12,14]. There have also been updates to the preferred imaging modalities used to assess possible renal artery stenosis in specific clinical circumstances[15]. These changes likely represent fewer stents being placed in patients who would not have benefited from stenting as well as more recent research which supports stenting in patients with more severe renovascular disease. Major studies since 2016 are summarized in Table 1 and discussed in the text as follows.

Table 1 Major studies on renal artery stenting since 2016.

Ref.	Study size and type	Baseline stenosis	Baseline blood pressure (mmHg) and anti-hypertensives	Baseline kidney function	Inclusion/exclusion criteria	Findings
Warfield et al[23], 2026 and Warfield et al[24], 2025	Single center, retrospective; 99 patients	85% patients with ≥ 70% and 60% of patients with ≥ 90% stenosis	165 ± 23/80 ± 14; 3.4 ± 1.3 medications	GFR: Mean = 39, SD = 24	Exclusion: FMD, transplant renal artery, ADPKD. Inclusion: Renal artery stent in at least one artery from January 2004 to December 2023	Renal artery stenting decreased both systolic and diastolic BP for five years. Defined daily dose and antihypertensive medication counts were decreased for 3 years and returned to baseline at 5 years post-stenting. Diastolic HTN (≥ 80 mmHg) and rapidly increasing blood pressure (≥ 15 mmHg increase over one year prior to stenting) were independently associated with larger decreases in systolic BP after stenting. ≥ 90% stenosis was a positive effect modifier for patients with only one of these predictors. Patients with neither predictor, one predictor, one predictor and ≥ 90% stenosis, or both predictors were progressively more likely to experience ≥ 20 mmHg decrease in systolic BP at 3 months post-stenting. GFR and serum creatinine increased from one year prior to stenting to time of stent placement and then decreased for five years after stenting in patients with rapidly declining kidney function (GFR decline of ≥ 5 mL/minute per 1.73 m2 within a year prior to stenting). This same pattern was seen in the overall cohort but was not significant. Both rapidly declining kidney function and ≥ 90% stenosis independently correlated with GFR improvement after stenting. Single functioning kidney was a positive effect modifier for patients with both predictors. Patients with neither predictor, one predictor, both predictors, or both predictors and single functioning kidney had progressively greater chance of ≥ 20% GFR increase following stenting
Modrall et al[33], 2017	Single center, retrospective; 61 patients	N/A	Median: 149/75; IQR: 140-179/67-89; on median 4 (IQR: 3-5) medications	GFR: Median = 34, IQR = 24-45	Serum creatinine > 1.5 at baseline	Patients after stenting were grouped as “responders” if eGFR increased by 20%. Responders had improved overall survival compared to non-responders. There were 17 responders and 44 non-responders of 61 included patients. The only significant independent predictor of being a responder was more rapid decline in kidney function. Responders had a median of -2.1% decline in GFR (IQR: -3.8% to -1%) per week whereas non-responders had median 0% (IQR: -0.8% to 1.4%) GFR decline per week prior to stenting
Modrall et al[30], 2023	Post-hoc analysis of CORAL trial	67.3% ± 11.4%	149.9 ± 23.2 on 2.1 medications	GFR: Mean = 58.0, SD = 23.4	CORAL patients	CORAL post-hoc analysis which showed that the stent plus medical therapy group had significantly more responders (classified as GFR improvement > 20% after stenting). Event free survival was significantly better for responders. However, only 25.6% of stent plus medical therapy patients were responders vs 17.1% of medical therapy patients being responders. Diabetes was a negative predictor of improved renal function after stenting and higher urine albumin creatinine ratio correlated with lower event free survival and worsened renal function
Modrall et al[22], 2020	Post-hoc analysis of CORAL trial	67.3% ± 11.4%	Systolic: Mean = 149.9, SD = 23.2; 2.1 medications	GFR: Mean = 58.0, SD = 23.4	CORAL patients	Patients were categorized as responders if their post-stenting BP was < 160/90 mmHg with fewer meds or diastolic < 90 mmHg. Predictors of this response were pre-stenting diastolic BP > 90 mmHg, clonidine use, and ≥ 4 anti-hypertensive medications. Responders and non-responders did not differ in event free survival. Percent of positive responders increased as the number of pre-procedural predictors increased
Modrall et al[32], 2023	VA database of 40 VA centers, retrospective; 695 patients	N/A	Median = 145/74, IQR = 133-161/67-81	GFR: Median = 46.0, IQR = 34.0-61.6	VA patients stented from 2000 to 2021	Overall, these VA patients had lower systolic and diastolic BP requiring the same number of medications with similar GFR. Responders (defined as GFR increase > 20% after stenting) had a 26.1% increase in GFR which was maintained throughout an average of 7.1 years of follow-up. Predictors of being a responder were CKD 3b and 4 (as opposed to stage 1, 2, 3a, and 5), more rapid decline in GFR, and diabetes (which had a negative correlation)
Courand et al[17], 2019	Single center, retrospective; 72 patients	78% ± 10%	157 ± 16/82 ± 10; 4.0 ± 1.0 medications	GFR: Mean = 52, SD = 11	Resistant HTN (SBP > 135 or DBP > 85 mmHg despite at least 3 antihypertensive drugs, including a diuretic)	Renal angioplasty (stenting in 66/72 patients) decreased daily ambulatory BP and anti-hypertensive burden but had no effect on GFR in patients with resistant HTN (defined as > 135 mmHg SBP or > 85 mmHg DBP despite 3 antihypertensive medications including one diuretic). High baseline SBP and low BMI predicted good response. These data were at first follow-up (mean = 57 days after stenting). Subset of patients at 1 year (n = 31) and 3 years (n = 18) showed similar results. Higher SBP, lower BMI, younger age, and higher GFR prior to stenting predicted lower post-stenting BP
Edgar et al[20], 2023	Three centers, retrospective; 72 patients	N/A	187 ± 30/89 ± 19	GFR: Median = 25, IQR = 8-42	All patients who had been stented	72 patients stratified by indication (HTN, renal dysfunction, or pulmonary edema). HTN and medication burden improved at one-year post-stenting without worsening kidney function. Intervention for kidney dysfunction led to improved GFR at 6-months post-stenting. Intervention for pulmonary edema was universally successful. Patients who would have been excluded from the CORAL trial but were included in this trial had greater reduction in serum creatinine and an equivalent reduction in SBP
Murphy et al[35], 2016	Post hoc analysis of CORAL trial	67.3% ± 11.4%	149.9 ± 23.2; 2.1 medications	GFR: Mean = 58.0, SD = 23.4	CORAL patients	Pre-stenting albuminuria < 22.5 mg/g (median baseline UACR of all CORAL patients) was associated with improved event free survival at five years in the stent plus medical therapy group as compared to the medical therapy group. This was true of the composite end point, cardiovascular disease related death, progressive renal insufficiency, and overall survival. Group with albuminuria was > 22.5 prior to randomization saw no difference in outcomes following either stenting plus medical therapy or medical therapy alone
Reinhard et al[16], 2022 (DAN-PTRA trial)	Two centers, prospective; 102 patients	40% of patients with ≥ 90% stenosis	166.2 ± 21.6/82.3 ± 12.3; 4 (3.7-4.2) antihypertensives	Median = 39.7, IQR = 23.5- 54.0	Stenosis > 70% plus resistant HTN (ambulatory SBP > 130 despite 3 meds 1 of which is diuretic), rapidly declining kidney function (reduction in GFR > 5 mL/minute per year), or recurrent heart failure (2 hospital admissions)/flash pulmonary edema (1 hospital admission) without another explanation such as left ventricular ejection fraction < 40%	Prospective study of patients with stenosis > 70% plus one or more of; resistant HTN (defined as SBP > 130 mmHg despite 3 antihypertensive medications, 1 of which is a diuretic), rapidly declining kidney function (defined as a reduction in GFR > 5 mL/minute per year), or recurrent heart failure (defined as at least 2 hospital admissions)/flash pulmonary edema (only 1 hospital admission) without another explanation such as a left ventricular ejection fraction < 40%. This study found lower BP and lower antihypertensive burden and higher GFR (in RDKF patients). UACR was also found to have increased over the year before stenting and then decreased over the year after stenting which was maintained for 3 years. High baseline ambulatory SBP, younger age, and recent angiotensin converting enzyme/angiotensin receptor blocker discontinuation predicted good a BP response. Female sex, high baseline ambulatory SBP, RDKF, recurrent heart failure/pulmonary edema, and stenosis > 90% predicted an improvement in GFR at 3 months
Vassallo et al[34], 2018	Singel center, retrospective; 263 patients	All patients with > 70% stenosis	Median = 155/80, IQR = 134-180/69-88	GFR: Median = 29.5, IQR = 20.4-41.3	≥ 70% unilateral or bilateral stenosis with or without high-risk features defined as: (1) FPE or acute decompensated HF without known left ventricular ejection fraction < 40%; (2) Systolic BP ≥ 160 and-or diastolic BP ≥ 100 despite three meds including a diuretic; and (3) rapidly declining kidney function with GFR slope < -3 over last year	Progression free survival from cardiovascular events, end stage renal disease, and these two combined with death was lower in in high-risk patients after stenting than after medical therapy alone. There was no difference in non-high-risk pts. This was driven by patients within the high-risk group who had RDKF (defined as a GFR slope < -3 mL/minute per year over last year. This was the 25th percentile of all patients pre-stenting). They also found that bilateral severe (> 70%). RAS was associated with increased risk for CVE, end stage renal disease, and both combined with death. Proteinuria < 1 g/day was associated with decreased risk of these
Li et al[26], 2026 (FAIR trial)	Single center, prospective; 101 patients	75.7% ± 14.3%	Median = 138/80, IQR = 125-146/74-89	GFR: Median = 55.8, IQR = 39.2-72.0	Home SBP ≥ 140 and/or DBP ≥ 90 as well as ≥ 50% angiographic stenosis	Patients were randomized to either an angiography guided group wherein all patients were stented, or a fractional flow reserve guided group wherein they were only stented if their FFR was < 0.8. Both groups had significantly reduced mean daily ambulatory SBP but required the same number of anti-hypertensives. When all stented patients with FFR < 0.8 were compared to stented patients with FFR > 0.8 and non-stented patients with FFR > 0.8 the mean daily ambulatory SBP was reduced in pts with FFR < 0.8 and these patients required fewer anti-hypertensive medications. FFR and percent stenosis angiographically were related logarithmically. FFR was measured using a dopamine induced gradient across the renal artery lesion
Iwashima et al[36], 2018	Single center, retrospective; 139 patients	60% of patients with severe (> 90% stenosis)	154 ± 25/78 ± 14	GFR: Mean = 45.2, SD = 20	SBP > 140 and/or diastolic > 90 or receiving antihypertensive treatment and stenosis > 50% on angiography along with uncontrolled HTN, declining kidney function, or cardiac destabilization syndrome	Patients were categorized by GFR, UACR or UPCR and assessed for their risk of meeting a composite end point (defined as all-cause death, myocardial infarction, stroke, adverse aortic events, or end-stage renal failure requiring regular hemodialysis). Normal to mild risk group: UACR < 3, UPCR < 15, GFR > 45. Moderate risk group: UACR = 3-30, UPCR = 15-50, GFR = 30-44. Severe risk group: UACR > 30, UPCR > 50, GFR < 30. GFR < 30 and severe UACR/UPCR independently predicted higher likelihood of meeting primary composite endpoint
Meredith et al[38], 2017	Single center retrospective; case-control; 188 patients	Survivors: 79% ± 10%; non-survivors: 82% ± 10%	Survivors: 156 ± 24/69 ± 13; non-survivors: 154 ± 28/68 ± 13	Survivors GFR: Mean = 79, SD = 73; non-survivors GFR: Mean = 59, SD = 29	≥ 70% atherosclerotic stenosis on concurrent coronary and renal angiography	Patients were split into survivors and non-survivors, and pre-stenting variables were compared between these two groups. Important predictors were identified using univariable regression and then included to a logistic regression. Previous MI, left ventricular ejection fraction < 35%, and GFR < 45 were predictors of mortality in a population of patients with renal artery stenosis > 70%. A survival benefit after RASt was seen in patients with 0 or 1 of these predictors but not 2 or 3 when compared to patients who were not stented. SBP, DBP, and stenosis severity were not correlated with survival
Koksal Cevher et al[21], 2026	Single center, retrospective; 69 patients	Left kidney: 85.5%; right kidney: 85.9%	169.8 ± 29.8/96.8 ± 16.6	Mean = 45.06, SD = 23.03	> 60% stenosis with either resistant HTN, acute kidney injury, or flash pulmonary edema	At one month follow up after stenting patients had lower SBP and DBP and required fewer antihypertensive medications, 55% of patients with acute kidney injury returned to baseline kidney function and 50% of patients with CKD exacerbations returned to baseline kidney function. Pulmonary edema resolved in all 16 patients it was found in at time of stenting
de Bhailis et al[19], 2024	Single center, retrospective; 127 patients	Median = 90%; IQR = 70%-95%	Median = 160/77, IQR = 139-181/66.5-99.5	Median = 40, IQR = 33.5-56	> 50% stenosis of at least one artery and uncontrolled hypertension, deteriorating kidney function, or heart failure syndrome	Of 127 patients were included in multidisciplinary discussions, 70 received medical therapy alone and 57 patients were stented (17 for HTN, 25 for deteriorating kidney function, 6 for heart failure syndromes, and 9 for severe anatomical stenosis); 82% of HTN patients experienced an improvement, 72% of deteriorating kidney function patients saw an attenuated GFR decline. All heart failure syndrome patients had no further heart failure admissions; 78% of anatomically severe stenosis patients had improved HTN and 55% had improved renal function. Additionally, patients who received stenting had more severe renal artery stenosis and more of them were on 3 or more anti-hypertensives prior to stenting compared to those who received medical treatment
Sens et al[18], 2019	Single center, retrospective; 49 patients	N/A	Stent group: 150 ± 31/85 ± 17; medical group: 142 ± 27/76 ± 15	Stent group GFR: Mean = 43, SD = 30; medical group GFR: Mean= 47, SD = 28	Atherosclerotic renal artery stenosis with peak systolic velocity > 180 cm/second on duplex ultrasound	Patients who were stented had lower SBP and DBP with fewer required anti-hypertensive medications and unchanged GFR at one-year post-stenting. Medical therapy patients had no changes in these parameters. Of note, the baseline DBP was higher in the stent group, and this was a small sample size (n = 23 in stent group and n = 26 in medical therapy group)
Herrmann et al[31], 2016	Single center, retrospective; 62 patients	Stenosis > 60% (stent group only)	Essential HTN group: 135 ± 19/71 ± 12; stent group: 143 ± 21/69 ± 7; medical group: 133 ± 19/67 ± 8	Essential HTN group GFR: Mean = 77, SD = 20; stent group GFR: Mean = 53, SD = 20; medical group GFR: Mean = 63.6, SD = 21.3	Non-diabetic white patients with renal artery stenosis > 60% compared to patients either with essential HTN or renal artery stenosis treated medically; 8 patients with bilateral high-grade stenosis excluded	After stenting the GFR of the previously stenotic kidney rose and had reduced biomarkers of hypoxia. However, the contralateral kidney experienced a drop in GFR. Overall, there was no change in GFR after stenting. These results must be interpreted with caution due to small sample sizes and many pre-stenting differences between groups. Particularly the differences in GFR prior to stenting
Ma et al[29], 2016	Meta-analysis of 7 studies; 253 patients	N/A	N/A	N/A	Patients with a single functioning kidney	Meta analysis of 7 papers including a total of 253 patients with single functioning kidneys, 77% of patients experienced improved or stabilized GFR after stenting. However, this should be interpreted with caution because the paper lacks significant detail about the other clinical and demographic information of these patients
Takahashi et al[37], 2020	Single center, retrospective; 398 patients	N/A	N/A	GFR: Mean = 42.5, SD = 17.8	All patients with data at pre-defined time points	398 patients who had GFR measured 6-12 months before stenting, immediately prior to stenting, and 6-12 months after stenting. Lower GFR post-intervention, diabetes, higher proteinuria (grouped per 100 g per day), and CKD stages 4 and 5 were associated with higher likelihood of renal replacement and all-cause mortality after stenting

GFR: Glomerular filtration rate; FMD: Fibromuscular dysplasia; ADPKD: Autosomal dominant polycystic kidney disease; BP: Blood pressure; HTN: Hypertension; eGFR: Estimated glomerular filtration rate; IQR: Interquartile range; N/A: Not applicable; VA: Veteran Affairs; CKD: Chronic kidney disease; SBP: Systolic blood pressure; DBP: Diastolic blood pressure; BMI: Body mass index; UACR: Urine albumin-to-creatinine ratio; CORAL: Cardiovascular Outcomes in Renal Atherosclerotic Lesions; RDKF: Rapidly declining kidney function; FPE: Flash pulmonary edema; HF: Heart failure; RAS: Renal artery stenosis; CVE: Cardiovascular event; FFR: Fractional flow reserve; UPCR: Urine protein creatinine ratio; RASt: Renal artery stenting; DAN-PTRA: Danish percutaneous transluminal renal angioplasty; PTRA: Percutaneous transluminal renal angioplasty.

STENTING FOR HYPERTENSION AND PREDICTORS OF BENEFIT

Many studies have shown a decrease in blood pressure following RASt[16-24]. However, the stenting group in CORAL only experienced approximately 2 mmHg lower SBP[9] compared to medical treatment, indicating that only patients with medically resistant hypertension should be considered for RASt. Currently stenting for hypertension is reserved for patients who have not seen sustained improvement in their blood pressure despite three medications including one diuretic[12,14].

The Danish PTRA (DAN-PTRA) trial[16] was a single arm prospective trial designed to assess the effects of RASt on hypertension, kidney function, and cardiac destabilization including recurrent heart failure and flash pulmonary edema (FPE) specifically in patients who met stenting guidelines as described above[14]. These patients also had severe anatomic disease with 40% of patients having stenosis > 90% and another 30% having stenosis between 80% and 90%. Of the 97 patients that received a stent in at least one renal artery, 96 met criteria for resistant hypertension with a pre-stenting average ambulatory blood pressure of 162/82 despite 4 antihypertensive medications. Blood pressure in these patients decreased for 2 years following stenting by approximately 25/9 mmHg. The defined daily dose of anti-hypertensive medications (which allows comparison across anti-hypertensive class) was reduced for the duration of the study as well. Courand et al[17] focused on 72 patients with ambulatory blood pressure ≥ 135/85 despite 3 anti-hypertensives including a diuretic and found reduced blood pressures and lower anti-hypertensive burden following RASt. Importantly, patients in this study also had high average ambulatory SBP of 157 despite 4 anti-hypertensives. Of note, this study did include 6 patients who received angioplasty without stenting and primary results were from a mean follow-up of 57 days after stenting. GFR was unchanged in this study.

Recently studies in 99 patients from Warfield et al[23] and Warfield et al[24] showed decreased systolic and diastolic blood pressure (DBP) for five years after stenting as well as reduced defined daily dose and medication count of antihypertensives for three years. The baseline SBP in this study was 165 mmHg despite 3.4 antihypertensives and was reduced by approximately 25 mmHg after stenting. Patients in this study were characterized by severe anatomic disease with approximately 60% of patients having ≥ 90% stenosis and 85% of patients having ≥ 70% stenosis. This study also conducted a subgroup analysis comparing patients who met criteria for medically resistant hypertension (SBP ≥ 140 mmHg despite three antihypertensives, including one diuretic) and those who did not. There was no difference in blood pressure across the five-year time frame, and both groups experienced significant decreases in systolic and DBP at all time points. This finding highlights the need for better predictive accuracy in patient selection which will be discussed below.

Two studies[18,19] analyzed the impact of multidisciplinary meetings to decide on stenting or medical therapy in patients with renal artery stenosis. In both studies, stented patients experienced lower blood pressures, but anti-hypertensive burden was reduced in one and stayed constant in the other. Sens et al[18] showed that patients who received medical management did not experience the same decrease in blood pressure as those who underwent stenting. GFR was unchanged in both groups. de Bhailis et al[19] also showed that 82% of patients who were stented primarily for hypertension had improved blood pressure control. Additionally, 72% of patients had improved GFR following stenting. These results indicated that better hypertension management may be achieved following stenting with multidisciplinary team involvement in patient selection. Of note, these studies were retrospective and the patients who were stented tended to have more severe disease as shown by significantly higher DBP in Sens et al[18] and significantly more severe renal artery stenosis in de Bhailis et al[19].

Other studies have analyzed patients based specifically on the indication for which stenting was performed[20,21]. Edgar et al[20] retrospectively analyzed 72 patients with severe hypertension (average of 187/89) across 3 sites. The 51 of these patients who were stented for hypertension experienced a decrease in SBP of 42 mmHg and required fewer anti-hypertensive medications without any significant change in kidney function. Koksal Cevher et al[21] achieved similar results showing lower SBP and DBP as well as fewer antihypertensive medications after 1 month follow up in 69 patients having received RASt.

PREDICTORS OF BENEFIT - RESISTANT HYPERTENSION

Finding variables that can predict lower blood pressure following RASt is important for both patient selection and for patient counseling. The DAN-PTRA trial[16] and Courand et al[17] both found that higher baseline ambulatory SBP predicted greater decrease in SBP. A post-hoc analysis of stented patients in the CORAL trial defined blood pressure response as post-stenting blood pressure < 160/90 on fewer anti-hypertensives or post-stenting DBP < 90. Patients with pre-stenting DBP > 90, requirement of ≥ 4 anti-hypertensives, or pre-stenting clonidine use were more likely to be a responder. Additionally, these predictors were additive meaning that patients who had more predictors were progressively more likely to be a responder. Interestingly these responders did not show a mortality benefit over the non-responders[22]. It should be noted as well that a post-stenting blood pressure of < 160/90 is not particularly exciting given that the average pre-stenting SBP in CORAL patients was approximately 150 mmHg[9]. Diastolic hypertension was also identified as predictive of SBP improvement in a prior meta-analysis which included eleven studies and 1552 patients[25].

Warfield et al[23] and Warfield et al[24] also found that higher DBP and rapidly increasing blood pressure (≥ 15 mmHg increase over one year prior to stenting) correlated with greater decreases in SBP following stenting. Importantly, these predictors were additive in that patients with more pre-stenting predictors present were progressively more likely to experience a decrease in SBP ≥ 20 mmHg, ≥ 90% angiographic stenosis was a positive effect modifier to delineate patients with one but not both predictors present.

Recently, results of the FAIR trial were published providing randomized prospective evidence for the use of fractional flow reserve (FFR) to guide decision making intra-procedurally[26]. In this trial 101 patients were randomized to either angiography guided RASt or FFR-guided RASt. Both groups saw equivalent reductions in blood pressure and antihypertensive burden. However, when patients were divided based on FFR, those with FFR < 0.8 had significantly improved blood pressure while requiring fewer anti-hypertensives whereas patients with FFR > 0.8 did not experience a benefit. The conclusion being that FFR should be employed intraprocedurally to guide whether a patient should be stented for hypertension. These results are further supported by prior retrospective studies showing dopamine induced trans-lesional systolic gradients of > 20 mmHg are correlated with greater blood pressure reduction[27,28]. While these results are exciting, it should be noted that FFR and stenosis severity (as graded by angiography) in this study follow a logarithmic relationship indicating that many of the patients found to have FFR < 0.8 are the same patients who had above approximately 80%-85% stenosis.

Many studies include different predictor variables which make comparison across studies difficult. However, the predictors which have been identified do point towards RASt in patients for hypertension only after having failed multiple medications and in the setting of a hemodynamically significant lesion. The best method by which a lesion is determined to be hemodynamically significant (percent stenosis, FFR, systolic gradients) remains unclear. Ideally RASt is reserved for a patient who has high SBP and high DBP despite at least three anti-hypertensives along with severe anatomic stenosis and reduced FFR. If a patient’s blood pressure has been increasing over the year prior to stenting this may also predict greater efficacy. Finding additional predictors of blood pressure improvement and better tying this to clinical improvement is an important area of future research. A flow diagram depicting a practical workflow for deciding which patients to stent for resistant hypertension is presented in Figure 1.

Figure 1 Flow diagram for deciding which patients to stent for resistant hypertension. When a patient remains hypertensive above 140 mmHg systolic and/or 90 mmHg diastolic despite three medications it is reasonable to get a renal artery duplex ultrasound. If this duplex ultrasound shows renal artery stenosis associated with elevated peak systolic velocity and/or elevated renal-aortic ratio then further history should be evaluated. This should include whether the patient has diastolic hypertension (> 80 mmHg) and/or rapidly increasing blood pressure (systolic blood pressure increase > 15 mmHg within a year). If both are present, then stenting is indicated. ¹These patients are highly likely to have ≥ 90% stenosis and/or fractional flow reserve < 0.8. If a patient has either rapidly increasing blood pressure or diastolic hypertension but not both then evaluation for ≥ 90% stenosis and/or fractional flow reserve < 0.8 using angiography is warranted, as is stenting if either are present. If stenosis is ≤ 90% and fractional flow reserve is > 0.8 then stenting is likely unnecessary. If a patient has neither diastolic hypertension or rapidly increasing blood pressure then renal angiography and stenting are likely unnecessary. RAS: Renal artery stenosis; PSV: Peak systolic velocity; RAR: Renal-aortic ratio; dHTN: Diastolic hypertension; RIBP: Rapidly increasing blood pressure; FFR: Fractional flow reserve.

STENTING FOR ISCHEMIC NEPHROPATHY AND PREDICTORS OF BENEFIT

RASt to improve renal function has seen mixed results across studies with heterogenous designs. Patients in the CORAL cohort had similar incidence of progressive renal insufficiency, renal replacement therapy, or death from a renal cause following stenting plus medical therapy or medical therapy alone[9]. Similarly, patients in the ASTRAL trial did not see a significant difference in the slope of GFR change following stenting plus medical therapy as compared to medical therapy alone[7]. A follow up study of the ASTRAL trial analyzed this cohort of patients in the context of the CORAL composite clinical outcome and found no benefit of stenting over medical therapy[8]. Contemporary studies have focused on higher risk patients than those which were included in the CORAL trial. Predictors of improved kidney function identified since 2016 are discussed below.

The DAN-PTRA study[16] investigated the impact of stenting on kidney function in patients with severe disease (as described above) with 63 patients meeting this study’s criteria for rapidly declining kidney function (RDKF) (GFR decline of > 5 mL/minute per 1.73 m² over one year). GFR in the overall cohort decreased prior to stenting and increased after stenting and this pattern was driven by the 63 patients with RDKF. Interestingly, urine albumin creatinine ratio (UACR) also followed this same pattern of increase prior to stenting and significant decrease after stenting. Both the effects on GFR and UACR remained significant for 2 years after stenting. Warfield et al[23] and Warfield et al[24] used the same definition of RDKF (GFR decline of > 5 mL/minute per 1.73 m² within one year prior to stenting) and found that GFR decreased for five years following stenting. In both studies, there was no improvement in GFR following RASt in patients without RDKF implying that stenting can salvage kidney function only while it is still declining but not once it has reached a steady state.

Improvements in kidney function following stenting were found in four other studies[19-21,29,30] with no change in kidney function found in an additional four studies[17,18,31,32]. Edgar et al[20] found that patients who would have been excluded from the CORAL trial experienced a significantly greater decrease in serum creatinine than those who would have been included. Separately, a post-hoc analysis of CORAL data[30] defined patients as “responders” if they experienced > 20% increase in their GFR and found that stented patients reached this threshold significantly more often than those receiving medical therapy only. Responders in this study also experienced improved event free survival. It should be noted however that only approximately 25% of stented CORAL patients experienced > 20% increase in GFR after stenting[30]. Herrmann et al[31] compared GFR in a small group (n = 10) of patients before and after stenting and found increased GFR in the ipsilateral kidney but reduced GFR in the contralateral kidney causing no change overall. In a multicenter review of 40 Veterans Affairs centers, Modrall et al[32] found no overall change in GFR following stenting but did find lower blood pressures and stable antihypertensive burden in these patients over 7.1 years of follow up.

Contemporary studies have also investigated the effects of stenting on clinical outcomes in the context of GFR changes[30,33,34]. In a single center retrospective review of 61 patients with serum creatinine > 1.5 mg/dL patients who experienced ≥ 20% improvement in GFR after stenting experienced a mortality benefit over those who did not[33]. This same group applied this analysis to the CORAL cohort and again found improved progression free survival from the composite CORAL endpoint in patients who experienced a > 20% improvement in GFR following stenting. Vassallo et al[34] retrospectively grouped 267 patients with ≥ 70% stenosis having received RASt into a high-risk group (n = 127) and a control group (n = 136). Progression free survival from cardiovascular events (CVE), end stage renal disease (ESRD), all-cause mortality, and these three combined were compared between stented and non-stented patients within the high-risk or control groups. High-risk features were FPE despite left ventricular ejection fraction (LVEF) > 40%, SBP > 160 mmHg or DBP > 100 mmHg despite three antihypertensives including a diuretic, and RDKF (≥ 3 mL/minute per 1.73 m² decrease in GFR within a year prior to stenting). High risk patients were found to have improved progression free survival from CVE, ESRD and the composite of these with all-cause mortality following stenting. No benefit was seen in the control group following stenting. Importantly, when high risk patients were stratified by indication for stenting it was found that patients stented for RDKF drove the improved progression free survival. Taken together, it seems that patients with severe anatomic disease paired with RDKF on average experience improved kidney function and clinical benefit from RASt. However, this is not true for patients with milder renovascular disease as was seen in the CORAL and ASTRAL trials.

PREDICTORS OF BENEFIT - ISCHEMIC NEPHROPATHY

In a post-hoc analysis of the CORAL trial, stented patients who had UACR lower than the median (< 22.5 mg/g) experienced improved event free survival in the composite CORAL endpoint as well as cardiovascular related death, progressive renal insufficiency and overall survival individually as compared to patients receiving medical therapy alone[35]. Additionally, there was no difference in blood pressure between the stent and medical therapy groups. The takeaway from this post-hoc analysis was that patients with low proteinuria prior to stenting are the most likely to receive clinical benefit. However, as the authors acknowledged, the 22.5 mg/g limit was arbitrary and is not necessarily the correct cut-off value for all patients. Indeed, other studies have found that lower levels of proteinuria correlate with better outcomes following RASt, although using different cut-off values[30,34,36,37]. Specifics of proteinuria analyses in these studies can be found in Table 1.

Another variable which has been shown to predict improved outcomes is acute or subacute worsening kidney function also termed RDKF. Six studies[16,23,24,32-34] found that RDKF correlated with improved outcomes following stenting while one study[37] found no association. In the DAN-PTRA trial it was found that the improved kidney function in the overall study sample was driven by patients with RDKF[16]. Warfield et al[23] and Warfield et al[24] found that improved GFR was exclusive to patients with RDKF. Similar results were found in a cohort of 695 Veterans Affairs patients where higher rates of GFR decline pre-stenting correlated with higher likelihood of GFR improvement after stenting[32]. Finally, Vassallo et al[34] found that patients being stented for RDKF were the ones driving improved progression free survival from CVE, ESRD and death in high-risk patients after RASt as compared to those managed medically.

Additional predictors studied include baseline kidney function, diabetes, and anatomic disease severity. Two studies found that higher pre-stenting GFR correlated with improved outcomes[17,37] and two found that lower pre-stenting GFR correlated with worse outcomes[36,38]. Two studies found that diabetes correlated with worse outcomes[22,32]. Finally, three studies[16,23,24] found that > 90% stenosis correlated with GFR improvement after stenting and another found that bilateral > 70% stenosis correlated with increased risk of CVE, ESRD and these plus all-cause mortality with an improvement in all three following RASt[34]. More predictors identified since 2016 are presented in Table 1.

Renovascular disease affecting the total functional mass of the kidney(s) may also predict salvage of renal function[12]. When one kidney is ischemic, the other will increase GFR to compensate[31] but if the entire functional renal mass is ischemic there can be no compensation. The DAN-PTRA trial[16] did not see a correlation between improved GFR and bilateral stenting. However, there were only 15 out of the 102 patients in the study that received a bilateral stent. Vassallo et al[34] showed that patients with bilateral ≥ 70% stenosis, especially combined with proteinuria < 1 g per day, experienced improved event free survival from a composite outcome similar to what was used in CORAL (discussed above). Warfield et al[23] and Warfield et al[24] showed that patients with RDKF and ≥ 90% stenosis who had a single functioning kidney experienced the largest improvement in GFR of any patient subgroup and that these patients had a 75% chance of ≥ 20% improvement in GFR. Importantly, patients with RDKF are the ones most likely to experience a mortality benefit following RASt[34] and this mortality benefit has been seen in patients who experience ≥ 20% improvement in GFR after RASt[30]. The findings of these studies illustrate the utility of RASt in patients with ischemic nephropathy affecting the entire functional renal mass, also termed global ischemia.

Improvement in kidney function following RASt seems to correlate strongly with clinical improvement. As such the variables which predict post-stenting improvement in kidney function also predict clinical improvement. These include RDKF, low proteinuria, absence of diabetes, severe anatomic stenosis, and single functioning kidney or bilaterally severe stenosis. The predictive value of these variables indicate that health of kidney parenchyma is needed if RASt is to be successful. This is also supported by older studies which identified renal resistance index (RI) (a proxy measure of fibrosis in the kidney) > 0.8 as predictive of worsened kidney function after stenting[39]. It should, however, be noted that RI may not be a reliable metric of kidney fibrosis or the health of the kidney parenchyma. In fact, studies of kidney tissue samples which directly measure the degree of kidney fibrosis and of microvascular disease have found no correlation between these metrics and RI. Therefore, the use of this metric clinically is shaky at best[40].

Stenting is beneficial in patients only once blood flow to the kidney is impaired to such a degree that GFR worsens. This aligns with the idea of RASt being most successful in kidneys which are subject to ischemia but in which the inflammatory cascade has not yet caused significant parenchymal fibrosis and resultant proteinuria[12]. The remaining challenge lies in finding methods to differentiate ischemic damage from inflammatory damage and fibrosis. One method to identify kidney ischemia could be the use of blood oxygen level dependent magnetic resonance imaging to find the fractional tissue hypoxia within a kidney[41]. This method provides a non-invasive way to evaluate kidney ischemia without the need for intravascular contrast agents. However, it has not been studied for the ability to predict improved kidney function following stenting. A flow diagram depicting a practical workflow for deciding which patients to stent for ischemic nephropathy is presented in Figure 2.

Figure 2 Flow diagram for deciding which patients to stent for ischemic nephropathy. When a patient is found to have worsening kidney function thought to be related to renal artery disease, the first step is a duplex ultrasound. If this shows elevated peak systolic velocities and renal-aortic ratios, the next step becomes evaluating the health of the kidney parenchyma. This is done by measuring either proteinuria or urine albumin to creatinine ratio. ¹It remains unclear what a good cutoff value for these lab tests is but urine albumin to creatinine ratio < 30 mg/g or proteinuria < 1 g per day would indicate healthy parenchyma. Next evaluate for rapidly declining kidney function (RDKF) with glomerular filtration rate trended over the last year. If RDKF is present, then renal angiography is indicated. Without RDKF, stenting is likely unnecessary. If angiography shows a renal artery stenosis ≥ 90% in a patient with a single functioning kidney (congenitally/surgically absent or atrophic contralateral kidney) or in bilateral renal arteries, then stenting is indicated. If the stenosis is ≤ 90% stenting is likely unnecessary. If there is a ≥ 90% stenosis unilaterally in a patient with two functioning kidneys then stenting can be performed but should involve the interventionalists’ clinical judgement about the individual case. RAS: Renal artery stenosis; PSV: Peak systolic velocity; RAR: Renal-aortic ratio; UACR: Urine albumin to creatinine ratio; RDKF: Rapidly declining kidney function; SFK: Single functioning kidney.

STENTING FOR FPE AND RECURRENT HEART FAILURE EXACERBATION

FPE and recurrent episodes of congestive heart failure are related to renal artery stenosis through multiple physiologic mechanisms including but not limited to the renin-angiotensin-aldosterone system and salt and water retention[12]. Hospitalizations for heart failure were included as part of the primary composite endpoint of the CORAL trial[9] and there was no difference in incidence between groups. However, patients were excluded from the CORAL trial if they had experienced a hospitalization for heart failure within the last 3 months[9,42]. Current guidelines recommend stenting in patients with hemodynamically significant renal artery stenosis and either recurrent heart failure exacerbations or FPE[14]. However, it does remain important to consider other causes of heart failure exacerbations particularly in a patient with reduced LVEF. This was illustrated by Meredith et al[38] who found that LVEF < 35% was independently associated with mortality in patients with renal artery stenosis > 70%.

Five studies investigated the effect of RASt on FPE and recurrent heart failure exacerbations[16,19-21,34]. The DAN-PTRA trial included 20 patients stented for FPE or recurrent heart failure exacerbations despite LVEF > 40%[16]. Three patients were not stented due to occlusive renovascular disease and 14 of the 17 stented patients (82%) had no further hospitalizations in the two years of follow up. de Bhailis et al[19] included 6 patients stented for FPE and found no future hospitalizations for heart failure in five years of follow-up. Similar results were found in Edgar et al[20] and Koksal Cevher et al[21] where patients stented for FPE experienced no future heart failure related hospitalizations over a follow up of 1-year and 1-month respectively.

These results highlight the efficacy of RASt in preventing future hospitalizations for properly selected patients presenting with FPE or recurrent heart failure exacerbations. Despite these results, it remains unclear whether RASt provides a survival benefit for patients with FPE. Vassallo et al[34] found that progression free survival from CVE, ESRD, and death in high-risk patients was improved after RASt as compared to medical therapy alone but that this was driven by patients presenting for RDKF (discussed above). This was not the case for patients who were stented for FPE despite LVEF > 40%[34]. Given these data, RASt likely leads to improved volume regulation but may not confer a survival benefit, meaning it may fill the role of diuretics in patients who are refractory.

Based on current studies there does seem to be a role for RASt in preventing future hospitalizations for patients presenting with FPE in the setting of a hemodynamically significant renal artery stenosis. However, ruling out other causes of heart failure exacerbation as well as demonstrating adequate left ventricular function prior to RASt is necessary. Future research investigating predictor variables of improved survival following RASt for FPE is also warranted and is currently scant.

RISKS OF RAST

The most common adverse events for RASt include arterial dissection, distal embolization or ipsilateral renal embolization, and renal artery occlusion/thrombosis. Many complications following RASt are minor and require minimal medical attention, such as an uncomplicated access site hematoma. Others may require a follow up minor procedure, such as a thrombin injection for an access site pseudoaneurysm. Some, however, do require an overnight stay in the intensive care unit, such as a vessel rupture and bleed requiring transfusion and/or embolization.

The CORAL cohort[9] had a complication rate of 5.25% (26/495 patients) in the stenting arm. In the ASTRAL trial[7], the periprocedural event rate (within 24 hours of the procedure) was higher at approximately 9% (31/359 patients). Of note, the ASTRAL trial included complications which were either not listed or did not occur in the CORAL trial, such as access site vessel damage, which contributed to the higher adverse event rate. In the DAN-PTRA trial[16] there were ten total complications occurring in 113 placed stents for a complication rate of 8.84%. In this trial the most common complication was access site (2 femoral artery and 2 brachial artery) pseudoaneurysm formation while only four patients experienced a dissection, vessel rupture, or thrombosis. The FAIR trial[26] reported a complication rate of 0.28% which included three total adverse events, one of which was an uncomplicated femoral hematoma. Retrospective studies reporting adverse event rates include Courand et al[17] (5.6%, 4/72 patients) and Warfield et al[24] (6.06%, 6/99 patients). Based on these studies it is reasonable to expect complications in approximately 5%-8% of patients with many complications being minor.

CONCLUSION

Following the negative results of the CORAL and ASTRAL trials, patient selection is critically important when considering RASt. Stenting for hypertension should be reserved for patients who have medically resistant hypertension despite three anti-hypertensives including a diuretic. Patients most likely to benefit are those who present with elevated SBP and DBP in the setting of anatomically significant (> 80%) stenosis. Recent work has also highlighted the role of FFR ≥ 0.8 in predicting patients most likely to experience reductions in blood pressure following stenting. Stenting for ischemic nephropathy is most effective in patients who have experienced a rapid decline in their kidney function paired with low levels of proteinuria although the proper threshold for these variables is not yet established. Care should be taken when considering RASt in patients with a history of diabetes or low baseline GFR (< 40 mL/minute/1.73 m²). Patients who experience improved kidney function after stenting are often the ones who experience improved clinical outcomes including reduced mortality. RASt for FPE or recurrent heart failure exacerbations is warranted when other causes have been ruled out, and the LVEF is > 40%. A practical flow diagram for clinical use in deciding which patients should receive RASt is presented in Figures 1 and 2.

Imaging and lab tests to assess a patient’s candidacy for RASt should include a duplex renal ultrasound to screen for hemodynamically significant stenosis and determine the ipsilateral RI, UACR to determine kidney parenchymal health, and serum creatinine to trend the patients GFR leading up to and after RASt. Other imaging which may be helpful includes computed tomography with and without contrast or magnetic resonance with and without contrast, however, care should be taken in the setting of a patient with reduced kidney function. Angiographically, patients must have a hemodynamically significant stenosis which is often defined as > 70%, however, more severe stenoses may respond better favoring a threshold of > 80% or even > 90%. Intravascular ultrasound can be used to better characterize the stenosis. Pressure gradients may also be helpful in determining whether a stenosis is hemodynamically significant especially in the setting of RASt for hypertension management.