Author + information
- Received December 1, 2008
- Revision received February 17, 2009
- Accepted March 3, 2009
- Published online June 23, 2009.
- Massoud A. Leesar, MD⁎,⁎ (, )
- Jai Varma, MD⁎,
- Adam Shapira, MD⁎,
- Ibrahim Fahsah, MD⁎,
- Seyed T. Raza, MD†,
- Ziad Elghoul, MD⁎,
- Anthony C. Leonard, PhD‡,
- Karthikeyan Meganathan, MS‡ and
- Sohail Ikram, MD⁎
- ↵⁎Reprint requests and correspondence:
Dr. Massoud A. Leesar, Division of Cardiology, University of Cincinnati, 231 Albert Sabin Way, MSB-3054, Cincinnati, Ohio 45267
Objectives We investigated the comparative accuracy of renal translesional pressure gradients (TPG), intravascular ultrasound (IVUS), and angiographic parameters in predicting hypertension improvement after stenting of renal artery stenosis (RAS).
Background The degree of RAS that justifies stenting is unknown.
Methods In 62 patients with RAS, TPG (resting and hyperemic systolic gradient [HSG], fractional flow reserve, and mean gradient) were measured by a pressure guidewire; IVUS and angiographic parameters (minimum lumen area and diameter, area stenosis, and diameter stenosis) were measured by quantitative analyses.
Results The HSG had a larger area under the curve than most other parameters and an HSG ≥21 mm Hg had the highest sensitivity, specificity, and accuracy (82%, 84%, and 84%, respectively) in predicting hypertension improvement after stenting of RAS. The average IVUS area stenosis was markedly greater in RAS with an HSG ≥21 mm Hg versus <21 mm Hg (78% vs. 38%, respectively; p < 0.001). After stenting, hypertension improved in 84% of patients with an HSG ≥21 mm Hg (n = 36) versus 36% of patients with an HSG <21 mm Hg (n = 26) at 12 months, p < 0.01; the number of antihypertensive medications was significantly lower in patients with an HSG ≥21 mm Hg versus <21 mm Hg (2.30 ± 0.90 vs. 3.40 ± 0.50, respectively; p < 0.01). By multivariable analysis, HSG was the only independent predictor of hypertension improvement (odds ratio: 1.39; 95% confidence interval: 1.05 to 1.65; p = 0.013).
Conclusions An HSG ≥21 mm Hg provided the highest accuracy in predicting hypertension improvement after stenting of RAS, suggesting that an HSG ≥21 mm Hg is indicative of significant RAS.
The discordance between high procedure success and moderate clinical response rates in patients with renal artery stenosis (RAS) may stem from the limitations of angiography for assessment of the significance of RAS. A number of series (1–4) demonstrated poor correlations comparing diameter stenosis by quantitative renal angiography with a number of the renal translesional pressure gradients (TPG), including resting systolic gradient (RSG), fractional flow reserve (FFR), hyperemic systolic gradient (HSG), and hyperemic mean gradient (HMG).
Both intravascular ultrasound (IVUS) and FFR are well-validated techniques for assessing the significance of a coronary artery stenosis (5–7). Recently, it was reported that renal FFR is a promising tool to identify hypertensive patients with RAS who would likely benefit from renal artery stenting (4). However, to date, no comparative studies of renal TPG, IVUS, and angiographic parameters have been reported to demonstrate which parameter(s) reliably predict hypertension improvement after stenting of RAS.
Accordingly, the objective of the present study was to compare the diagnostic accuracy of renal TPG, IVUS, and angiographic parameters in predicting hypertension improvement after stenting of RAS.
From December 2004 to August 2006, 62 patients with hypertension and unilateral RAS (50% to 90% diameter stenosis by visual estimation) underwent assessment of atherosclerotic ostial RAS at our center. Exclusion criteria were severe renal dysfunction as evidenced by serum creatinine ≥3.0 mg/dl or kidney length <8.0 cm, and presence of accessory renal arteries. The institutional review board approved the protocol and all patients gave written informed consent.
Blood pressure and renal function measurements
Blood pressure was measured according to guidelines published by the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure VII (8). Blood pressure was obtained from patients seated in a chair after at least 5 min of rest; it was measured twice, and the average of the 2 blood pressure values was recorded. Hypertension was defined as systolic blood pressure ≥140 mm Hg and/or diastolic blood pressure ≥90 mm Hg. Patients with accelerated or refractory hypertension on 2 or 3 antihypertensive medications, respectively, were enrolled into the study. Improvement in hypertension was defined as diastolic blood pressure <90 mm Hg and/or systolic blood pressure <140 mm Hg or a reduction in diastolic blood pressure by at least 15 mm Hg with the same or reduced number of antihypertensive medications. These definitions are in accordance with guidelines for reporting of renal artery revascularization in clinical trials (9).
Before discharge, antihypertensive medications were adjusted according to the following algorithm: the first-line therapy included an angiotensin-converting enzyme inhibitor, angiotensin receptor blocker alone, or combined with a thiazide diuretic; the second-line therapy included long-acting calcium antagonists, beta-blockers, alpha-blockers, combined alpha- and beta-blockers, and hydralazine. Patients were followed by the investigators 3, 6, and 12 months after the procedure. At each visit, in addition to blood tests, blood pressure was measured and surveillance of antihypertensive medications was performed. If blood pressure was not at the goal, the doses of antihypertensive medications from the first-line therapy were increased or an antihypertensive agent from the second-line therapy was added and then recommendations were made to treating physicians to titrate the doses of antihypertensive medications or to add another antihypertensive agent from the second-line therapy to achieve the target blood pressure of ≤140/190 mm Hg in patients without comorbidities and ≤130/80 mm Hg in patients with diabetes and/or kidney disease, as recommended by the Joint National Committee VII guidelines (8).
After renal angiography, a 0.014-inch pressure guidewire (RADI Medical Systems, Uppsala, Sweden) was advanced into the renal artery through the guiding catheter. After equalization of pressures, the pressure transducer was advanced through the stenosis and RSG was measured. Next, a 30-mg bolus dose of papaverine was administered directly into the renal artery to induce hyperemia, as previously reported (2–4). After papaverine injection, the guiding catheter was retracted from the ostium of the renal artery to prevent dampening of pressures, and then HSG, FFR, and HMG were measured.
After obtaining pressure measurements, an IVUS catheter (Atlantis SR Pro, Boston Scientific, Natick, Massachusetts) was advanced into the renal artery over pressure guidewire. Ultrasound images were recorded after initiation of automated pullback at a speed of 0.5 mm/s, starting approximately 10 mm distal to the lesion. A SuperVHS videotape was used to record all studies for offline analysis.
All patients underwent renal pressure measurements and an IVUS procedure. After performing IVUS, all patients underwent renal artery angioplasty followed by stenting using the standard technique. Inclusion of patients for the study was based on the visual estimation of stenosis (RAS with a diameter stenosis of 50% to 90%). Stent size was determined by IVUS media-to-media diameter measured at a normal-looking segment distal to post-stenotic dilation; stent length was also determined by IVUS. Embolic protection devices were not used in any of patients. Angiographic success was defined as a <30% residual stenosis after stenting (9).
All IVUS images were analyzed off-line by an analyst blinded to the pressure measurements using computerized planimetry (TapeMeasure, INDEC Systems, Mountain View, California), according to previously validated and published protocols (3,12,18).
Quantitative renal angiography
Quantitative renal angiography was performed off-line by a skilled analyst blinded to the results of IVUS and pressure measurements using validated, automated, edge-detection software (QCA-CMS 5.2 system, Medis Medical Imaging Systems, Raleigh, North Carolina), according to previously validated and published protocols (3,6,10,11). A representative example of quantitative renal angiography, IVUS analysis, and renal pressure measurements is shown in Figure 1.
Continuous variables were analyzed using paired or unpaired Student ttests; categorical variables were compared using the chi-square test. A receiver-operating characteristic (ROC) analysis was performed to determine the optimal cutoff values of TPG, IVUS parameters, and angiographic parameters (as continuous variables) in predicting hypertension improvement (as a dichotomous variable) at 12 months. The cut-points selected were those that yielded the greatest sum of sensitivity and specificity. The area under the curves (AUCs) resulting from ROC analyses were compared using the method suggested by DeLong et al. (12). Univariate predictors of hypertension improvement with p values of <0.05 were entered simultaneously into a multivariable logistic model. Logistic regression results are presented as odds ratio with 95% confidence intervals. Means are presented with ± SD. A probability value <0.05 (2-sided) was considered statistically significant.
Baseline characteristics of 62 patients are summarized in Table 1.The angiographic and procedural success rates were 100%. After stenting, 2 patients developed femoral artery pseudoaneurysm, which was treated by direct thrombin injection.
Predictors of blood pressure improvement after stenting of RAS
The ROC analyses of parameters, including renal TPG (RSG, HSG, FFR, and HMG), IVUS parameters (MLA, MLD, area stenosis, and plaque plus media area), angiographic parameters (MLD and diameter stenosis), and clinical parameters (systolic, diastolic, and mean blood pressure) are shown in Table 2.HSG measured by the pressure guidewire had a larger AUC than most of the other parameters by the ROC analysis in predicting hypertension improvement at 12 months (Table 2, Figs. 2Ato 2D). The AUC for HSG was significantly greater than the AUCs for parameters such as MLD and plaque plus media by IVUS; MLD and diameter stenosis by angiography; and systolic, diastolic, and mean blood pressure in predicting hypertension improvement at 12 months, p < 0.05 (Table 2). In addition, the AUCs for FFR and MLA by IVUS were significantly greater than the AUCs for MLD by angiography, systolic, diastolic, and mean blood pressure in predicting hypertension improvement at 12 months, p < 0.05 (Table 2).
Furthermore, of the HSG values, an HSG ≥21 mm Hg provided the highest sum of sensitivity and specificity (sensitivity 82%, specificity 84%) and the highest accuracy (84%) in predicting hypertension improvement after stenting of RAS (Table 2). Likewise, in order to identify the independent predictors of hypertension improvement, we first assessed the parameters individually, as shown in Table 3.In the univariate model, renal TPG (RSG, HSG, FFR, and HMG); IVUS parameters (MLA, MLD, area stenosis, and plaque plus media area); and MLD by angiography were significantly associated with hypertension improvement (Table 3). When the above-mentioned parameters were placed simultaneously in a multivariable logistic model, HSG was the only parameter that predicted blood pressure improvement independent of the other predictors in the model (odds ratio: 1.32; 95% confidence interval: 1.05 to 1.65; p = 0.013).
Comparative data of IVUS, angiographic, and hemodynamic parameters
Having shown that an HSG ≥21 mm Hg provided the highest accuracy in predicting hypertension improvement, we next compared IVUS, angiographic, and hemodynamic parameters among patients with an HSG ≥21 mm Hg versus an HSG <21 mm Hg (Table 4).HSG was ≥21 mm Hg in 36 patients with RAS (58%); HSG was <21 mm Hg in 26 patients (42%). IVUS MLA and IVUS MLD were significantly smaller, but IVUS area stenosis and IVUS plaque plus media area were significantly greater in patients with an HSG ≥21 mm Hg versus an HSG <21 mm Hg. Of the angiographic parameters, only MLD was significantly smaller in patients with an HSG ≥21 mm Hg versus an HSG <21 mm Hg. RSG and HMG were significantly higher, but FFR was significantly lower in patients with an HSG ≥21 mm Hg versus an HSG <21 mm Hg.
There has been significant variability in the association of the preceding parameters in predicting major adverse outcomes such as death, myocardial infarction, stroke, or progression of renal insufficiency; thus, the clinical outcome of the present study is limited to the incidence of death, hypertension improvement, and serum creatinine levels at follow-up. During the 12-month follow-up period, no patient died or was otherwise lost to follow-up; 3 patients (1 patient with an HSG ≥21 mm Hg and 2 patients with an HSG <21 mm Hg) developed accelerated hypertension and underwent repeat renal angiography to assess for possible in-stent restenosis. Angiography demonstrated minimal in-stent restenosis. In these patients, hypertension was then adequately controlled by increasing the dose of angiotensin-converting enzyme inhibitors and thiazides.
After stenting, systolic and diastolic blood pressures, at 3-, 6-, and 12-month follow-up, were significantly lower in patients with an HSG ≥21 mm Hg (n = 36) than in those with an HSG <21 mm Hg (n =26) (Figs. 3Aand 3B). At 3- and 6-month follow-up, 89% and 86% of patients who had an HSG ≥21 mm Hg met the criteria for hypertension improvement compared with 38% and 42% of patients in whom HSG was <21 mm Hg, respectively; p < 0.01 (Fig. 3C). At 12 months, hypertension improvement was sustained in 84% of patients with an HSG ≥21 mm Hg (n = 36) compared with 36% of patients with an HSG <21 mm Hg (n = 26); p < 0.01 (Fig. 3C).
Before stenting, the number of antihypertensive medications was not significantly different between the groups (3.03 ± 0.69 vs. 2.92 ± 0.42). After stenting, the number of antihypertensive medications at 6- and 12-month follow-up was significantly lower in patients with an HSG ≥21 mm Hg than in those with an HSG <21 mm Hg (2.30 ± 0.54 vs. 3.40 ± 0.50 and 2.30 ± 90 vs. 3.40 ± 0.50, respectively; p < 0.01) (Fig. 4A).In patients with an HSG ≥21 mm Hg, the doses of antihypertensive medications were either significantly lower or remained unchanged compared with baseline at 12-month follow-up (Table 5).In contrast, among patients with an HSG <21 mm Hg, doses of the majority of antihypertensive medications were significantly higher compared with baseline at 12-month follow-up (Table 5). Furthermore, the doses of the majority of antihypertensive medications were significantly lower in patients with an HSG ≥21 mm Hg compared with those with an HSG <21 mm Hg at 12-month follow-up (Table 5).
Serum creatinine levels were not significantly different at baseline, at 6-, and 12-month follow-up between the groups (1.22 ± 0.45 mg/dl vs. 1.15 ± 0.40 mg/dl, 1.18 ± 0.45 mg/dl vs. 1.10 ± 0.25 mg/dl, and 1.05 ± 0.30 mg/dl vs. 1.15 ± 0.35 mg/dl) (Fig. 4B).
To the best of our knowledge, the present study is the first prospective study to compare the diagnostic accuracy of renal TPG, IVUS, and quantitative renal angiography in predicting hypertension improvement after stenting of RAS. Our data demonstrated that an HSG ≥21 mm Hg measured by pressure guidewire is the strongest predictor of hypertension improvement after stenting of RAS, suggesting that an HSG ≥21 mm Hg is indicative of hemodynamically significant RAS. In contrast, diameter stenosis measured by quantitative renal angiography did not predict blood pressure improvement. At 12-month follow-up, blood pressure, doses, and number of antihypertensive medications were significantly lower in patients with an HSG ≥21 mm Hg than in those with an HSG <21 mm Hg.
Assessment of renal artery stenosis
Guidelines for renal artery revascularization (13) suggested that a significant RAS is defined as the presence of 50% to 70% diameter stenosis by visual estimation, with a peak translesional gradient of at least 20 mm Hg, or a mean gradient of at least 10 mm Hg measured with a ≤5-F catheter or pressure guidewire. However, the use of a 4- or 5-F catheter would overestimate the pressure gradient. In this respect, Colyer et al. (1) reported that a 4-F catheter significantly overestimated the severity of RAS because a 0.014-inch pressure guidewire compared with a 4-F catheter would occupy 6% versus 24% of the renal artery, respectively.
In the coronary circulation, maximal hyperemia is essential in determining the physiological significance of stenoses detected by angiography. Likewise, it has been demonstrated that a vasodilator reserve exists in the renal circulation. By analogy, it is conceivable to induce hyperemic gradient by vasoactive agents to assess the significance of RAS. In this respect, Beregi et al. (14) showed that intrarenal injection of isosorbide dinitrate or papaverine significantly increased renal blood flow in pigs, but the hyperemia was significantly greater with papaverine compared with isosorbide dinitrate because papaverine dilates small resistance vessels; isosorbide dinitrate dilates only epicardial vessels. In the present study, the use of papaverine was based on previous studies (2–4,14–16) in which intrarenal papaverine was used extensively to either induce hyperemia or to assess the significance of RAS.
In 3 recent series, the pressure guidewire was used to determine the significance of RAS. Jones et al. (17) measured HSG in 22 patients with RAS; these investigators demonstrated that, after stenting of RAS in 13 patients with HSG >20 mm Hg, systolic blood pressure significantly improved at follow-up. The present study has validated the findings of Jones et al. (17) that HSG is a significant predictor of hypertension improvement after stenting of RAS. De Bruyne et al. (18) demonstrated a Pd/Paratio (the ratio of distal renal pressure to aortic pressure) of 0.90 as a threshold for renin release. These investigators (18) measured Pd/Paratio by a pressure guidewire while a balloon catheter was inflated inside of the stented segment of renal artery to induce a controlled pressure gradient between the aorta and distal renal artery. They concluded that a Pd/Paratio <0.90 could be considered a hemodynamically significant RAS; however, this index needs to be validated with clinical outcomes. Mitchell et al. (4) reported that stenting in 17 patients with RAS resulted in a significant hypertension improvement in patients who had a renal FFR <0.80 compared with those with an FFR >0.80. It is worth noting that Mitchell et al. (4) did not measure HSG to assess hypertension improvement after stenting of RAS. In addition, our data demonstrated that renal FFR was not an independent predictor of hypertension improvement after stenting of RAS. Although the measurement of FFR is useful in the coronary circulation (10), a large body of evidence supports the low predictive power of FFR in predicting hypertension improvement (11,15,16). An explanation is probably linked to lower vasodilator reserve in the renal circulation than in the coronary microvasculature, and a number of investigators have shown renal flow reserve to be variable and in some cases is very low (5,15,16).
With respect to the measurement of renin, a number of series (19,20) have demonstrated that renal vein renin activity increases in patients with renal artery stenosis. Although the renal vein renin assay is possibly beneficial, however, the measurements are impractical to obtain.
Blood pressure improvement after stenting of renal artery stenosis
Despite a high procedural success rate of renal artery stenting, an improvement in hypertension has been inconsistent. This most likely reflects the absence of predictors for blood pressure improvement after renal artery stenting. Rocha-Singh et al. (21) reported that stenting of RAS based on visual estimation resulted in blood pressure improvement in 47% of their patients at 24 months. Others reported an improvement in blood pressure in 62% (22) and 76% of patients at 12 months (23). These studies required more stringent criteria before stenting, including duplex ultrasound or evidence for a critical RAS based on quantitative renal angiography. Because the correlation between renal pressure measurements and angiographic diameter stenosis is poor, it is likely that patients without significant pressure gradient have been included in these trials. Such an inconsistent blood pressure response to renal stenting underscores the value of renal pressure measurements for appropriate patient selection. The CORAL (Cardiovascular Outcomes in Renal Atherosclerotic Lesions) trial (24) randomized patients based on severity of angiographic stenosis. These data suggest that hemodynamic significance is much more important than angiographic severity. As such, stenting of hemodynamically nonsignificant lesions in the CORAL trial is unlikely to confer any benefit compared with medical therapy. Instead, it may be more useful to compare repair of hemodynamically significant lesions with medical therapy. In particular, in light of the observed 19.7% major adverse event rate that has been reported after renal artery stenting (21), HSG-guided renal artery stenting allows for better selection of patients who may benefit from renal artery stenting. The limitation of our study is inherent to small sample size; a large randomized trial comparing outcomes among HSG-guided stenting versus medical therapy is warranted.
In this first report of the comparative accuracy of the renal TPG, IVUS, and angiography in patients with RAS, we demonstrated that in patients with RAS, an HSG ≥21 mm Hg was an independent predictor of blood pressure improvement after stenting of RAS. Our results showed that among patients with RAS, regardless of their angiographic severity, an HSG ≥21 mm Hg indicates a hemodynamically significant RAS. Furthermore, in this setting, HSG can be easily measured after renal angiography with a pressure guidewire, circumventing the need for IVUS or renal vein renin study to determine the significance of RAS. This would, in turn, facilitate decision-making regarding medical therapy versus stenting in patients with RAS.
Continuing Medical Education (CME) is available for this article.
- Abbreviations and Acronyms
- area under the curve
- fractional flow reserve
- hyperemic mean gradient
- hyperemic systolic gradient
- intravascular ultrasound
- minimum lumen area
- minimum lumen diameter
- renal artery stenosis
- receiver-operating characteristic
- resting systolic gradient
- translesional pressure gradients
- Received December 1, 2008.
- Revision received February 17, 2009.
- Accepted March 3, 2009.
- American College of Cardiology Foundation
- Kern M.J.,
- Meier B.
- Leesar M.A.,
- Abdul-Baki T.,
- Akkus N.I.,
- Sharma A.,
- Kannan T.,
- Bolli R.
- Fearon W.F.,
- Luna J.,
- Samady H.,
- et al.
- Jasti V.,
- Ivan E.,
- Yalamanchili V.,
- Wongpraparut N.,
- Leesar M.A.
- Varma J.,
- Shapira A.,
- Ikram S.,
- Leesar M.A.
- Hirsch A.T.,
- Haskal Z.J.,
- Hertzer N.R.,
- et al.
- Manoharan G.,
- Pijls N.H.,
- Lameire N.,
- et al.
- De Bruyne B.,
- Manoharan G.,
- Pijls N.H.,
- et al.
- Rocha-Singh K.,
- Jaff M.R.,
- Rosenfield K.
- Zeller T.,
- Frank U.,
- Muller C.,
- et al.