Author + information
- Received October 19, 2007
- Revision received March 12, 2008
- Accepted March 31, 2008
- Published online June 1, 2008.
- Ehtisham Mahmud, MD, FACC⁎ (, )
- Thomas W.R. Smith, MD,
- Vachaspathi Palakodeti, MD, FACC,
- Owais Zaidi, MD,
- Lawrence Ang, BS,
- C. Robinson Mitchell, MD,
- Nayab Zafar, MD,
- Guilherme Bromberg-Marin, MD,
- Shahin Keramati, MD, FACC and
- Sotirios Tsimikas, MD, FACC
- ↵⁎Reprint requests and correspondence:
Dr. Ehtisham Mahmud, Cardiovascular Catheterization Laboratories, University of California, San Diego Medical Center, 200 West Arbor Drive, San Diego, California 92103-8784.
Objectives This study sought to identify angiographic parameters of favorable clinical response to renal artery stenting.
Background Stenting improves blood pressure (BP) control in patients with renal artery stenosis (RAS), but markers predicting a favorable clinical response are limited.
Methods Renal perfusion was quantified in hypertensive patients (BP ≥140/90 mm Hg) without RAS by determining renal frame count (RFC) (angiographic frames [30 frames/s] for contrast to reach distal renal parenchyma after initial renal artery opacification) and renal blush grade (RBG) (0: none, 1: minimal, 2: normal, 3: hyperemic parenchymal blush). It was hypothesized that stenting unilateral RAS in hypertensive patients would result in decreased RFC and increased RBG, which might predict BP reduction.
Results The RFC in 17 consecutive hypertensive patients without RAS (control group) (64.4 ± 14.2 years, 12 male, 22 kidneys) was 20.1 ± 5.4, whereas RBG was 2.33 ± 0.66. In 24 consecutive hypertensive patients with unilateral RAS (study group) (72.7 ± 11.3 years, 8 male), reduced RFC (26.6 ± 9.1 to 21.4 ± 6.7, p < 0.001) and increased RBG (1.63 ± 0.71 to 2.13 ± 0.85, p = 0.03) were observed after renal stenting. At 6 months, reduced BP (systolic BP 150.6 ± 15.6 mm Hg to 128.6 ± 15.5 mm Hg, p < 0.001; diastolic BP 77.2 ± 15.6 mm Hg to 68.3 ± 10.4 mm Hg, p = 0.022) without change in number of hypertensive medications was observed. Clinical responders (systolic BP reduction >15 mm Hg) had a greater decrease in RFC (7.7 ± 4.6 vs. 1.7 ± 5.1, p = 0.009) and 78.6% of patients with >4 RFC decrease were responders (p = 0.024).
Conclusions This study shows that quantitative indices of renal perfusion (RFC and RBG) are impaired in patients with RAS and improve after stenting, and that RFC reduction is associated with BP reduction.
Atherosclerotic renal artery stenosis (RAS) and fibromuscular dysplasia are the most common treatable causes of secondary hypertension and are associated with both renovascular hypertension and ischemic nephropathy (1,2). Atherosclerosis is a common, progressive vascular disease, particularly in patients with diabetes and multiple cardiovascular disease risk factors, and atherosclerotic RAS is the most common cause of secondary hypertension (1,3). Furthermore, in patients undergoing coronary angiography who are not suspected of having RAS, the incidence of significant RAS is 14.3% in the presence of severe hypertension, unexplained renal dysfunction, acute pulmonary edema with hypertension, or severe atherosclerosis (4). Endovascular renal artery (RA) stenting is associated with a reduction in blood pressure (BP) and use of a lower number of antihypertensive medications (5–7). However, hypertensive patients with RAS may also have a component of essential hypertension because hypertension usually persists in the majority of patients despite successful percutaneous revascularization with balloon angioplasty and stenting (5). In the ASPIRE-2 (A Study to evaluate the safety and effectiveness of the Palmaz balloon expandable stent in the REnal artery after failed angioplasty) study, hypertensive patients with atherosclerotic RAS were treated with endovascular stents, but only half of the study cohort was noted to have a BP reduction in response to stenting (5). Nevertheless, RA stenting has become increasingly common for managing hypertension, preventing progressive renal insufficiency, and better treating congestive heart failure and angina refractory to medical management.
Although RA stenting is increasingly being used, there have been limited biochemical and ultrasound parameters helpful in identifying patients likely to respond to renal revascularization therapy (8,9). Recently, renal fractional flow reserve <0.8 has been noted to predict a clinical response (BP reduction) after renal stenting (10,11). However, no other angiographic parameters of renal flow identify patients who may respond to RA stenting. Mulumudi et al. (12) assigned a continuous and quantitative schema to the rate of renal perfusion using the frame counter on standard angiographic images and measuring the number of frames required for dye to reach specific anatomical landmarks in the distal renal parenchyma after initial RA opacification. Using this method, they reported a range of values for the renal frame count (RFC) in a small number of normal subjects and patients with fibromuscular dysplasia. However, it was not determined whether the RFC correlated with clinical benefit of RA stenting.
The RFC predominantly reveals the rate of flow in the larger renal arterial vessels, but may not adequately assess microvascular perfusion. Therefore, the aim of this study was to determine RFC and quantify renal blush grade (RBG) as a parameter of microvascular perfusion in subjects with hypertension without RAS. Further, we hypothesized that in subjects with unilateral RAS, improvement in these 2 variables of renal perfusion would predict clinical benefit after RA stenting.
The study protocol was approved by the University of California, San Diego Human Subjects Protection Program. Hypertensive patients (BP ≥140/90 mm Hg) despite treatment with 2 or more antihypertensive medications undergoing coronary angiography were consented for a screening distal abdominal aortogram with possible selective renal angiography. The control group consisted of a consecutive series of 17 such patients (22 kidneys) with serum creatinine <2.0 mg/dl and evidence of atherosclerotic renal disease who underwent selective RA angiography, and had no pressure gradient demonstrable with a 5-F end hole catheter. The RFC and RBG were measured in these subjects. The study group comprised of 24 subjects undergoing unilateral RA stenting. The RA angiograms for all patients who underwent successful stenting for atherosclerotic RAS over a 2-year period were reviewed for inclusion. We excluded patients with angiographic images inadequate for RFC or RBG measurement, patients who had undergone bilateral RA stenting, patients with a creatinine >2.0 mg/dl, and patients who did not have a 3- to 6-month follow-up clinical examination available. Furthermore, any patients who had undergone RA stenting for a kidney <8 cm or angiographically evident pruning of the distal renal microvasculature before stenting also were excluded.
Demographics and clinical outcomes
Baseline demographic data were obtained from a prospectively collected outcomes database and reconfirmed by a chart review in all patients. The BP measurements were averaged for the 2 clinic visits before and 2 visits after the RA stent procedure provided these were obtained within 3 months before and within 6 months after the procedure. All BP measurements were obtained in the clinic in a uniform manner after a patient had been in the room resting for 10 min and using an arm cuff with the patient in the sitting position. Clinical responders to RA stenting were defined as subjects who had a systolic BP reduction >15 mm Hg (approximately 10% reduction in systolic BP) using the same or lower number of medications at 6-month clinical follow-up.
For the control group, selective renal angiography was performed in both left anterior oblique and right anterior oblique views with hand injections using 5-F catheters. The angles were set up to image the RA, branches, and kidney in the most perpendicular dimension to the image intensifier for measuring RFC and RBG. For the study group, renal stenting was performed using standard techniques, generally using a 6-F guiding catheter, in a similar manner to optimally image the RAs and kidney of interest.
All images were digitally obtained on the Heartlab (AGFA, Mortsel, Belgium) imaging system. The RFC measurement was adapted from the method described by Mulumudi et al. (12). Briefly, the number of cineangiographic frames taken for contrast dye to reach from the proximal RA (contrast filling the transverse diameter of the artery) to the distal landmark of the smallest cortical branch with images obtained at 30 frames/s were measured (Fig. 1). The RBG was adapted from the similar criteria used for the myocardial blush grade by van't Hof et al. (13) and evaluated after 2 s of initial arterial opacification (i.e., at the 60th frame after opacification). It was defined semiquantitatively as grade 0 = no parenchymal blush/contrast opacification of cortical vessels or inability of contrast clearance from renal parenchyma after initial opacification; grade 1 = minimal parenchymal blush/contrast opacification of cortical vessels; grade 2 = complete parenchymal blush/contrast opacification of cortical vessels; grade 3 = hyperemic parenchymal blush/brisk clearance of contrast from cortical vessels (Fig. 2). Quantitative RA angiographic parameters were determined using the guide catheter as a reference (QVA-CMS, version 5.2, MEDIS Medical Imaging Systems, Leiden, the Netherlands). Two independent operators who were blinded to the demographics and outcomes data interpreted all of the angiograms and made measurements. Reproducibility of RFC and RBG was assessed by repeating these measurements at an interval >3 months.
Data were analyzed using SPSS software (version 14.0, SPSS Inc., Chicago, Illinois). Data are reported as mean ± standard deviation (SD). Comparative analysis in the study group was performed with paired t test analysis for continuous variables and McNemar chi-square analysis for categorical variables. Comparative analysis for the clinical responders and nonresponders was performed with unpaired t test analysis for continuous variables and chi-square analysis for categorical variables. Fisher exact test was used when the expected value in any cell was ≤5. To determine reproducibility of RFC and RBG, a reliability analysis was performed and intraclass correlation coefficients were determined. A value of p < 0.05 was deemed to be statistically significant.
The control group consisted of a consecutive series of 17 hypertensive patients (64.4 ± 14.2 years, 12 male, 22 kidneys) without RAS. They had preserved renal function (creatinine 1.23 ± 0.55 mg/dl and creatinine clearance 76.9 ± 26.7 ml/min), and their RFC was 20.1 ± 5.4 without a difference between the right and left RAs. The RBG for the group was 2.33 ± 0.66 (Fig. 3). The reliability coefficient as defined by intraclass correlation coefficient for RFC was 0.991 (95% confidence interval 0.984 to 0.995, p < 0.001) and for RBG 0.924 (95% confidence interval 0.864 to 0.957, p < 0.001).
During the study period, 51 patients who underwent RA stenting were evaluated. A total of 27 patients were excluded (5 inadequate angiographic images, 5 bilateral renal stenting, 11 incomplete clinical follow-up available, 5 with creatinine >2.0 mg/dl, 1 kidney was <8.0 cm) and the remaining 24 patients formed the study group. This group was elderly (72.7 ± 11.3 years), predominantly female (67%), with a high percentage of patients having multiple cardiovascular disease risk factors, and documented atherosclerotic disease in other vascular territories (Table 1). There was an equal distribution of right and left RAs being stented (12 each) with 20 ostial lesions, 3 proximal RA lesions, and 1 patient with an ostial and midvessel lesion requiring treatment. In addition, 3 patients had stent placement in a diseased accessory RA in the same kidney. There were a total of 28 stents placed for an average of 1.17 stents/kidney with 27 stents being Herculink (Guidant, Temecula, California), and 1 stent a Taxus (Boston Scientific, Natick, Massachusetts) drug-eluting stent placed in an accessory RA.
Angiographic analysis showed significant reduction in diameter stenosis in the study group after RA stenting (71.3 ± 9.8% to 2.9 ± 10.5%, p < 0.001). This was accompanied by a reduction in the RFC (26.6 ± 9.1 to 21.4 ± 6.7, p < 0.001) (Fig. 3A) and an increase in RBG (1.63 ± 0.71 to 2.13 ± 0.85, p = 0.03) (Table 2,Figs. 3B and 4). The reliability coefficients for RFC and RBG for the study group both before and after renal stenting were similar to those of the control group (data not shown). Clinical follow-up showed a significant reduction in BP (systolic BP 150.6 ± 15.6 mm Hg to 128.6 ± 15.5 mm Hg, p < 0.001; diastolic BP 77.2 ± 15.6 mm Hg to 68.3 ± 10.4 mm Hg, p = 0.022; mean arterial pressure 101.6 ± 14.1 mm Hg to 88.4 ± 10.1 mm Hg, p < 0.001), and a trend toward the use of a lower number of antihypertensive medications without any adverse effect on renal function (Table 3,Fig. 5). Patients were further stratified as clinical responders (systolic BP reduction >15 mm Hg) or nonresponders (systolic BP reduction ≤15 mm Hg) to RA stenting at 6-month follow-up. Based on this parameter, 58% (14 of 24) of the patients had a significant reduction in their BP, whereas 42% (10 of 24) did not. The clinical responders had a greater decrease in BP (systolic BP 155.9 ± 17.2 mm Hg to 121.9 ± 15.3 mm Hg, p < 0.001; diastolic BP 82.6 ± 16.6 mm Hg to 68.1 ± 11.9 mm Hg, p = 0.007), whereas the nonresponders only had a minimal decrease in their BP (systolic BP 143.1 ± 9.3 mm Hg to 138.0 ± 10.2 mm Hg, p = 0.018; diastolic BP 69.5 ± 10.4 mm Hg to 68.6 ± 8.5, p = 0.86). The clinical responders had a higher baseline RFC (28.8 ± 9.2 vs. 23.6 ± 8.4, p = 0.17), a similar baseline RBG (1.64 ± 0.74 vs. 1.60 ± 0.70, p = 0.89), and a greater reduction in the RFC after stenting (ΔRFC 7.7 ± 4.6 vs. 1.7 ± 5.1, p = 0.009) (Fig. 6A), but a similar increase in RBG (ΔRBG 0.43 ± 0.64 vs. 0.60 ± 0.52, p = 0.48). Among patients with a baseline RFC ≥25 (n = 12), 75% were responders to RA stenting (p = 0.11). In the group of patients with a >4 RFC decrease (n = 14), 78.6% were responders to RA stenting, compared with only 30% in those with ≤4 RFC decrease (p = 0.024) (Fig. 6B).
This study shows that patients with atherosclerotic RAS and hypertension have decreased renal perfusion as measured by angiographic parameters of RFC and RBG. After RA stenting, these renal perfusion indices return to levels comparable to a control population with hypertension but without RAS. Furthermore, this study shows for the first time that these quantitative measures of renal perfusion predict a strong likelihood of clinical response after RA stenting. Quantitative measures of renal perfusion such as the ones described in this report may allow more precise diagnosis of renovascular hypertension and response to interventional therapies.
Progressive atherosclerosis of the RAs is a result of the effects of multiple factors on the vasculature, including chronic hypertension, diabetes, hypercholesterolemia, and smoking. Narrowing of the large and/or small arteries of the kidneys leads to renal ischemia, and affected areas of the kidney respond to progressively lower blood flow by producing renin leading to elevated levels of angiotensin II and aldosterone (14). These hormones lead to exacerbation of systemic hypertension, aggravating the atherosclerotic process and completing a positive feedback loop. Higher BP further damages the arteries of the kidneys, causing renal ischemia, fibrosis, and progressive loss of renal mass and renal function, which is also a stimulus for higher BP. Whereas chronically elevated systemic BP eventually leads to small vessel sclerosis, small arterioles in a kidney perfused by a stenotic artery are initially protected from such adaptive changes to pressure. However, chronic hypoperfusion caused by a stenosis can also lead to progressive atrophy and interstitial fibrosis. Up to this point, there have not been any accurate angiographic or noninvasive methods to determine whether patients have either viable renal parenchyma responsive to reperfusion or atrophy beyond recovery. Angiographic measurements of RFC and RBG are objective measures of renal perfusion from the large RA to the small cortical vessels and are easy to measure with a high degree of reproducibility. Further, we show that improvement in these parameters is associated with a reduction in BP after renal stenting.
The patients who had a response to renal stenting with lowering of BP had the highest baseline RFC and lowest RBG, and also had significant reduction in their RFC after stenting. This might represent a group of patients with hemodynamically significant stenosis who have recruitable functional renal reserve. Although it did not reach statistical significance (p = 0.11), a baseline RFC ≥25 was 75% predictive of responsiveness to renal stent therapy. We chose a strict criteria of at least a 10% reduction in systolic BP (>15 mm Hg) at follow-up, and with a lower threshold used to judge the success of renal stenting, a greater percentage of patients would be deemed responders. Our goal was, however, to define a point at which the measured clinical improvement would be unequivocal especially for the treatment of unilateral RAS. A larger study population should help define the optimal baseline RFC most predictive of clinical response.
The majority of patients with a >4 RFC decrease after stenting responded to renal stenting with improved BP control. The patients in whom RFC fails to improve or worsens may have renal parenchymal fibrotic changes that are irreversible or might have undergone distal embolization of atherosclerotic material during the procedure. Henry et al. (15) have shown that with the use of distal embolic protection devices, renal stenting leads to a lower likelihood of renal function deterioration, which may be caused by lower distal embolization. This has not been proven clearly, and measurement of RFC in future clinical trials may help to determine the role of embolic protection devices in improving or preventing deterioration in renal perfusion.
The measures of successful RA revascularization include better BP control, decreased episodes of angina and congestive heart failure, and improvement of renal function (16). Noninvasive evaluation of the renal parenchyma with ultrasonography and Doppler assessment to measure the renal resistive index (a measure of microvascular renal flow) is useful in stratifying responders to renal revascularization, with patients having a resistive index of <0.7 more likely to respond to revascularization (9,17). Noninvasive measures of renal blood flow by magnetic resonance imaging can also predict clinical outcomes in patients undergoing RA revascularization for RAS (18). Silva et al. (8) have noted that in similar patients, serum brain naturetic peptide >80 pg/ml is predictive of a response to RA stenting, whereas patients with a brain naturetic peptide level lower than this had no change in their BP after stenting. We show that quantifying renal perfusion by measuring RFC and RBG may be another tool that can help predict the presence of significant RAS and success of renal stenting. If validated in further studies, these angiographic parameters may be of immediate clinical utility in the angiographic laboratory by enabling the identification of patients likely to benefit from renal stenting. Furthermore, they may be useful as surrogate angiographic markers of clinical benefit in clinical trials in a manner analogous to coronary reperfusion studies.
This study is limited by the retrospective analysis of the study group and the relatively small sample size, which precluded the ability to further differentiate whether RFC and RBG provide similar or complementary information. Adjunctive simultaneous measures of renal flow, including ultrasound assessment or renal fractional flow reserve measurement, also were not performed, and therefore correlation with other measures of flow could not be made. The role of hemodynamic changes, including variation in cardiac output and systolic BP and their potential effects on RFC and RBG, also were not determined. Absence of distal embolic protection device use by our group precluded the ability to evaluate the effect of these devices on RFC and RBG. Finally, the assessment of RBG is semiquantitative and awaits validation with other measure of microvascular renal perfusion.
Quantitative measures of renal perfusion (RFC and RBG) are impaired in the presence of RAS and improve to normal levels immediately after RA stenting. This reduction in RFC after stenting is associated with BP reduction and >4 RFC reduction after stenting predicts BP reduction in 78% of patients. This study provides tools for evaluating renal perfusion and determining potential efficacy of RA stenting in the catheterization laboratory.
Dr. Mahmud received research grant support from Guidant and Boston Scientific Corporations. Drs. Zafar and Bromberg-Marin have received training grants in interventional cardiology from Guidant, Cordis, and Boston Scientific Corporations.
- Abbreviations and Acronyms
- blood pressure
- renal artery/arteries
- renal artery stenosis
- renal blush grade
- renal frame count
- Received October 19, 2007.
- Revision received March 12, 2008.
- Accepted March 31, 2008.
- American College of Cardiology Foundation
- Buller C.E.,
- Nogareda J.G.,
- Ramanathan K.,
- et al.
- Rocha-Singh K.,
- Jaff M.R.,
- Rosenfield K.,
- ASPIRE-2 trial investigators
- Dorros G.,
- Jaff M.,
- Mathiak L.,
- et al.
- Silva J.A.,
- Chan A.W.,
- White C.J.,
- et al.
- Manoharan G.,
- Pijls N.H.,
- Lameire N.,
- et al.
- van't Hof A.W.,
- Liem A.,
- Suryapranata H.,
- et al.
- White C.J.