Genitourinary Imaging
Detection of Renal Artery Stenosis
Prospective Comparison of Captopril-Enhanced Doppler Sonography, Captopril-Enhanced Scintigraphy, and MR Angiography
Audio Available | 
Share
OBJECTIVE. The objective of our study was to compare the value of captopril-enhanced Doppler sonography, captopril-enhanced renal scintigraphy, and gadolinium-enhanced MR angiography for detecting renal artery stenosis.
SUBJECTS AND METHODS. Forty-one patients with suspected renovascular hypertension were prospectively examined with captopril-enhanced Doppler sonography, captopril-enhanced renal scintigraphy, gadolinium-enhanced MR angiography, and catheter angiography. The sensitivity and specificity of each technique for detecting renal artery stenosis measuring 50% or greater and 70% or greater were compared using the McNemar test. Positive and negative predictive values were estimated for populations with 5% and 30% prevalence of renal artery stenosis. Kappa values for interobserver agreement were assessed for both gadolinium-enhanced MR angiography and catheter angiography.
RESULTS. For detecting renal artery stenosis measuring 50% or greater, the sensitivity of gadolinium-enhanced MR angiography (96.6%) was greater than that of captopril-enhanced Doppler sonography (69%, p = 0.005) and captopril-enhanced renal scintigraphy (41.4%, p = 0.001). No significant difference in specificity was observed among modalities. For renal artery stenosis measuring 50% or greater, positive and negative predictive values were respectively 62% and 86% for captopril-enhanced Doppler sonography, 49% and 76% for captopril-enhanced renal scintigraphy, and 53% and 98% for gadolinium-enhanced MR angiography. Interobserver agreement was high for both gadolinium-enhanced MR angiography (κ = 0.829) and catheter angiography (κ = 0.729).
CONCLUSION. Gadolinium-enhanced MR angiography is the most accurate noninvasive modality for detecting renal artery stenosis greater than or equal to 50%. The use of captopril-enhanced Doppler sonography in combination with gadolinium-enhanced MR angiography for identifying renal artery stenosis needs to be evaluated with a cost-effectiveness analysis.
| Introduction |   |
|---|
Renal artery stenosis is the leading cause of curable hypertension. Estimates suggest that the prevalence of renovascular disease as a cause of hypertension ranges from 0.5% to 5% in the general population [1, 2] to as high as 45% in selected patients with suggestive clinical features [3]. Catheter angiography, which is accepted as the gold standard for the detection of renal artery stenosis, is not an ideal screening method because it is invasive and expensive. Catheter angiography requires administration of iodinated contrast material and exposure to ionizing radiation. A reliable noninvasive diagnostic test is needed to select patients for invasive diagnostic and therapeutic approaches.
During the last decades, several noninvasive imaging modalities have been evaluated for their ability to detect renal artery stenosis. Renal scintigraphy [4,5,6,7,8,9,10] and Doppler sonography [11,12,13,14,15] that show captopril-induced changes provide indirect evidence of the presence of renal artery stenosis and have proven helpful in screening patients with this condition. However, data concerning the reliability of these techniques are inconsistent and vary among studies. Many authors have reported disappointing results for both techniques [7, 16,17,18,19,20,21,22]. More recently, substantial advances have been achieved with gadolinium-enhanced three-dimensional MR angiography for the identification of renal artery stenosis [23,24,25,26,27]. Yet the value of noninvasive modalities for the detection of renal artery stenosis has not been sufficiently defined, and the most useful diagnostic strategy remains undetermined. Furthermore, the performance of a given screening modality can be influenced by the prevalence of true renovascular disease in the population studied. Thus, prospective comparisons are needed to evaluate the value of each modality and to identify a clear diagnostic strategy. Our study was designed to compare the accuracy of captopril-enhanced renal scintigraphy, captopril-enhanced Doppler sonography and gadolinium-enhanced MR angiography for the detection of renal artery stenosis in patients with clinically suspected renovascular hypertension, and to determine the predictive value of these methods for identifying renal artery stenosis both in nonselected populations and in populations selected on the basis of clinical features.
| Subjects and Methods | |
|---|
Between January 1998 and May 1999, 41 patients (15 men, 26 women; age range, 41-78 years; mean, 64 years) were prospectively enrolled in this study. Patient selection was based on the presence of one or several of the following clinical features: onset of hypertension before the age of 25 years or after 45 years; severe hypertension (malignant hypertension, grade III or IV retinopathy, hypertensive encephalopathy, or diastolic blood pressure > 115 mm Hg); refractory hypertension (systolic blood pressure > 160 mm Hg or diastolic blood pressure > 95 mm Hg despite optimal doses of three antihypertensive drugs); acceleration of hypertension by more than 15% within the preceding 6 months; or abdominal or flank bruit. During this time, 30 patients were not included in the study for one or more of the following reasons: creatinine clearance of less than 40 mL/min; hyperkalemia (potassium > 5.5 mmol/L), because of the risk of nephrotoxicity induced by iodinated contrast material; history of stroke or transient ischemic attack with a carotid bruit, because of the risk of hypotension induced by captopril; history of allergy to angiotensin-converting enzyme inhibitors or iodinated contrast material; or contraindications to MR imaging (e.g., pacemaker, ocular metallic foreign bodies). Thirty-six additional patients refused to undergo all examinations, sometimes because they had undergone one or more noninvasive studies with a normal result. The mean arterial systolic over diastolic blood pressure of the study population was 162 ± 23 over 85 ± 12 mm Hg and the mean creatinine clearance level was 107 ± 38 mL/min. Forty of the study patients had two kidneys and one had a solitary kidney, for a total of 81 kidneys studied.
All patients underwent intrarenal Doppler sonography before and after captopril administration, captopril-enhanced scintigraphy, gadolinium-enhanced MR angiography, and catheter angiography within a 3-month period. The sequence of examinations depended on the accessibility of the imaging modality at the time of imaging. Angiotensin-converting enzyme inhibitors and calcium blockers were discontinued 2-5 days (depending on the half-life of the medication) before radionuclide and Doppler sonographic examinations [10]. No surgery or endovascular procedure was done between any imaging modalities.
The study was approved by our institutional ethics and research committees, and written informed consent was obtained from all patients.
Doppler sonographic examinations were performed with a Spectra unit (Diasonics, Milpitas, CA) equipped with a 3.5-MHz phased array transducer. After we identified intrarenal arteries with color-flow imaging using a posterior oblique approach, spectral velocity waveforms were obtained at an angle of insonation of less than 60° from segmental arteries at the superior, mid (anterior and posterior), and inferior portions of the kidney. We used the smallest velocity scale, the lowest wall filter, and a sweep time of 2 sec. Each patient underwent two Doppler sonographic examinations using the same technique: the first, baseline examination was followed by a second examination performed 1 hr after the oral administration of 25 mg of captopril, in keeping with the recommendations of the consensus report on angiotensin-converting enzyme inhibitors for detecting renovascular hypertension [10]. We used a pattern recognition approach according to previously published criteria [15]. The most abnormal Doppler spectrum (provided that it was reproducible) was selected by the investigator for each kidney before and after the administration of captopril and was morphologically classified into one of the three types described by Oliva et al. [15]. Type A represents a normal spectrum with an early systolic peak and a steep linear early systolic rise. Type B includes a normal spectrum without an early systolic peak but with a steep linear early systolic rise. Type C represents abnormal spectrum with a decrease of the early systolic rise. A type C Doppler spectrum was considered indicative of renal artery stenosis. Additional measurements of the resistive index, acceleration, and acceleration time of early systolic rise were obtained for the selected Doppler spectrum. Acceleration and acceleration time thresholds for positive results were set at 390 cm/sec2 and 0.06 sec, respectively, for the baseline examination and at 440 cm/sec2 and 0.09 sec for the captopril-enhanced examination [15]. For each kidney, Doppler findings were considered positive if either the quantitative (acceleration, time of acceleration) or morphologic (pattern recognition) evaluation was abnormal. In cases of disagreement between quantitative and qualitative (pattern recognition) evaluations, assessment of renal artery patency was based on pattern recognition, given the high interobserver correlation previously established with this method (κ = 0.95) [15]. Direct Doppler imaging of the proximal renal arteries could not be performed consistently because of technical limitations. Therefore, only intrarenal Doppler sonography was used for this comparative study in order to minimize the number of exclusions. The investigator's level of confidence in the captopril-enhanced Doppler sonography interpretation was rated on a five-point scale as follows: very high, high, fair, low, and very low.
All examinations were performed and analyzed by one of two investigators who were unaware of the findings of the other techniques.
Baseline and captopril-enhanced 99mTc-mercaptoacetyltriglycine (99mTc-MAG3) scintigraphy was performed in all patients using a 1-day, 25-mg captopril protocol as recommended by the Working Party Group on Determining the Radionuclide of Choice [28]. 99mTc-MAG3 is a protein-bound radiopharmaceutical tracer, and its clearance is almost exclusively through tubular secretion. 99mTc-MAG3 was preferred to other tracers because of the high extraction efficiencies, its image quality, and its favorable dosimetry. Patients were instructed to be well hydrated before the examination. Because chronic administration of angiotensin-converting enzyme inhibitors may reduce the sensitivity of scintigraphy, this medication was withheld from all patients for 2-5 days before the examination and was replaced by other drugs when indicated. A large-field-of-view gamma camera interfaced with a computer was positioned beneath the patient to obtain standard posterior views of the kidneys. Images were stored in a 64 × 64 word-mode pixel matrix. Time—activity curves were generated. Data were obtained for a minimum of 30 min. After the baseline study, an oral dose of 25 mg of captopril was administered, and the patient was instructed to drink 300-500 mL of water. The patient was then placed in the supine position and blood pressure was monitored at frequent intervals. The captopril-enhanced study was initiated 60 min after captopril administration. Results were interpreted according to the guidelines of the Society of Nuclear Medicine [29]. The most important criterion for detecting renal artery stenosis was unilateral parenchymal retention of the radiopharmaceutical after captopril administration. A change in the 20-min to maximum uptake ratio of 0.15 or greater, an increased delay of 2 min before maximum uptake, or changes superior or equal to 2 in the renogram grade (from a 5-level scale) were considered indicative of renal artery stenosis. Patients with abnormal baseline findings indicative of reduced renal function that were not modified after captopril administration were considered to have an intermediate probability of renal artery stenosis.
All examinations were performed and interpreted by one investigator who was unaware of the findings of the other imaging studies.
MR angiography was performed with 1.5-T unit (Magnetom Vision; Siemens, Erlangen, Germany) using a phased array body coil. Examination of renal arteries consisted of two sequences in the coronal plane: before and during a dynamic IV administration of 0.2 mmol/kg of body weight of gadopentetate dimeglumine (Magnevist; Berlex Canada, Montreal, Canada) to provide background subtraction and to increase vessel-to-background contrast. Images were acquired with the following parameters in a single breath-hold: three-dimensional gradient-echo technique; TR/TE, 3.4/1.4 msec; receiver bandwidth, 890 Hz per pixel; field of view, 300 × 300 mm2; matrix, 128 × 256; volume coverage, 100 mm; slice thickness, 1.8 mm (after interpolation, the effective thickness was 0.9 mm); scanning time, 25 sec. The contrast material was administered at a rate of 1.5-2 mL/sec. The transit time of the contrast material was determined using a test-bolus sequence of 2 mL and dynamic region-of-interest analysis of the signal intensity at the level of the renal arteries. Image acquisition was started 2 sec before the signal intensity peak. In all patients, T1- and T2-weighted fast spin-echo sequences were obtained before gadolinium-enhanced MR angiography to evaluate the morphologic status of the kidneys, such as the parenchymal volume and signal. Maximum-intensity-projection reconstructions and multiplanar reformations were processed after subtraction.
Gadolinium-enhanced MR angiography examinations were reviewed by two independent investigators without knowledge of the results of any other examination. Renal angiograms were graded for image quality using a three-point scale: optimal, when a high degree of contrast enhancement was obtained without motion artifacts; suboptimal, when the quality was sufficient for analysis of the main renal arteries but without a high degree of contrast enhancement; and inconclusive, when poor opacification or major motion artifacts were observed. Combined analysis of source images, maximum-intensity-projection reconstructions, and multiplanar reformations was used to analyze renal arteries and to quantify stenosis. The percentage of stenosis was calculated using a precision caliper, a magnifying lens, and the following formula: (D — d) / D × 100, where D is the diameter of the uninvolved segment of renal artery and d is the diameter of the stenotic segment. In cases of multiple renal arteries, the most stenotic artery was considered. When more than one stenosis were identified in a single renal artery, the most severe stenosis was used for analysis. In cases of intravascular signal void, renal artery stenosis was considered greater than 70%.
Catheter angiography was performed on a digital subtraction system (DFP 2000; Toshiba Medical System, Otawara-Shi, Japan) through the femoral approach in all patients using a 5-French pigtail catheter introduced with the Seldinger technique. A standard posteroanterior projection of the abdominal aorta was obtained in all patients using 40 mL of 32% iodinated contrast material (Visipaque 320; Nycomed Imaging, Ontario, Canada) injected at a rate of 20 mL/sec. Additional projections and selective angiograms were obtained if necessary at the discretion of the investigator.
All angiograms were reviewed independently by the same two investigators who reviewed gadolinium-enhanced MR angiographic examinations. A minimum delay of 1 month was observed between gadolinium-enhanced MR angiography and catheter angiography interpretation sessions. The identification of patients was concealed to avoid bias resulting from patient recognition. The catheter angiography interpretation session was done after the gadolinium-enhanced MR angiography interpretation session. Investigators analyzed image quality and measured renal artery stenosis in the same manner as for gadolinium-enhanced MR angiography. To categorize kidneys and patients with catheter angiography, two thresholds—50% and 70%—were used to define renal artery stenosis. Discrepancies among investigators that led to the classification of renal artery stenosis into different categories at catheter angiography were resolved by consensual interpretation to establish the standard of reference.
Kidneys in which at least one examination was inconclusive were excluded from the comparative analysis. According to the findings in each modality, kidneys and patients were classified using a two-point scale as follows: absence of renal artery stenosis, or presence of renal artery stenosis. Catheter angiography was considered the standard of reference. Indeterminate results with captopril-enhanced Doppler sonography or captopril-enhanced scintigraphy were considered positive to facilitate statistical analysis. This attitude is also in keeping with our usual clinical practice.
The sensitivity and specificity for renal artery stenosis detection were calculated for each technique on the basis of the findings at catheter angiography using 50% and 70% thresholds for renal artery stenosis. The McNemar test was used to compare the obtained values. For gadolinium-enhanced MR angiography, the results of the first investigator were used for the comparative analysis.
The predictive value of each technique for detecting renal artery stenosis greater than or equal to 50% was estimated using Bayesian analysis for a nonselected population (with a 5% renal artery stenosis prevalence [30]) and for a population selected on the basis of clinical criteria (the prevalence of renal artery stenosis was set at 30% according to previously published data [1]).
The interobserver variability for interpreting gadolinium-enhanced MR angiography and catheter angiography was assessed using the kappa value and intraclass correlation coefficient. On the basis of the kappa value, agreement was defined as follows: poor, less than 0.20; fair, 0.21-0.40; moderate, 0.41-0.60; good, 0.61-0.800; and excellent, 0.80-1.00. A 95% confidence interval (CI) was assigned to the calculated kappa value. The degrees of stenosis measured with gadolinium-enhanced MR angiography and with catheter angiography were compared in kidneys with renal artery stenosis greater than or equal to 50% using the Student's t test.
Statistical analysis was performed with a statistical software system (SAS for Windows, version 6.12; SAS Institute, Cary, NC). Differences were considered statistically significant when p values were less than 0.05.
| Results | |
|---|
Catheter angiography was considered optimal in 98% (investigator 1, 99%; investigator 2, 97%) of kidneys and suboptimal in 2% (investigator 1, 1%; investigator 2, 3%). Ninety renal arteries, including nine accessory or multiple arteries, were identified in the 81 kidneys. Intrarenal artery Doppler waveforms were obtained in all kidneys studied. Doppler examinations were scored with a high or very high level of confidence in 66 kidneys (81%). Only two examinations (2%) had a low or very low level of confidence. Captopril-enhanced renal scintigraphic examinations were available and were considered diagnostic in all kidneys. Gadolinium-enhanced MR angiography examinations were considered optimal in 72% of kidneys (investigator 1, 75%; investigator 2, 69%) and suboptimal in 25% (investigator 1, 22%; investigator 2, 28%). In one patient, both the left and the right gadolinium-enhanced MR angiography renal angiograms (2%) were considered inconclusive and were excluded from statistical analysis. Therefore, 79 kidneys were available for the comparative study. All but three accessory renal arteries identified at catheter angiography were visualized on gadolinium-enhanced MR angiography.
Catheter angiography revealed the presence of renal artery stenosis greater than or equal to 50% in 31 (76%) of 41 patients and in 41 (52%) of 79 kidneys. The population with renal artery stenosis (mean degree of stenosis, 68 ± 11%) consisted of 12 men and 19 women having a mean age of 65 ± 9 years and a mean arterial systolic over diastolic blood pressure of 163 ± 22 over 84 ± 12 mm Hg. Twenty-seven patients had atherosclerotic lesions, and four patients had fibromuscular dysplasia. Three kidneys had a totally occluded renal artery. The nonstenotic population (mean degree of stenosis, 20 ± 19) consisted of three men and seven women having a mean age of 60 ± 11 years and a mean arterial systolic over diastolic blood pressure of 160 ± 27 over 87 ± 17 mm Hg.
The mean standard deviation between the two investigators for the degree of stenosis determined with catheter angiography was 9% (range, 0-41%). Concordance between investigators in quantifying the degree of stenosis was excellent, with an intraclass correlation coefficient of 0.90. When kidneys were categorized using the two-point ordinal scale with a 50% threshold, disagreement occurred in 11 kidneys, with a kappa value of 0.729 (range, 0.581-0.877). When the threshold was set at 70%, agreement remained good despite a slight decrease of the kappa value to 0.691 (range, 0.495-0.887).
Agreement between investigators for renal artery stenosis quantification with gadolinium-enhanced MR angiography (intraclass correlation coefficient = 0.88) was similar to that observed with catheter angiography (intraclass correlation coefficient = 0.90). The mean standard deviation between the two interpreters of gadolinium-enhanced MR angiography for the degree of stenosis was 10% (range, 0-74%). Agreement between investigators measured with kappa coefficients calculated with 95% CIs for identifying renal artery stenosis measuring 50% or greater was excellent for gadolinium-enhanced MR angiography (κ = 0.829 [95% CI, 0.699-0.959]) and good for catheter angiography (κ = 0.0.729 [95% CI, 0.581-0.877]). However, kappa values were slightly lower for the 70% threshold than for the 50% threshold. With the 70% threshold, kappa values were considered good (κ = 0.691 [95% CI, 0.495-0.887]) for catheter angiography and moderate (κ = 0.592 [95% CI, 0.385-0.799]) for gadolinium-enhanced MR angiography. Compared with catheter angiography, gadolinium-enhanced MR angiography overestimated the degree of stenosis (Fig. 1A,1B,1C). Among 41 kidneys with renal artery stenosis greater than or equal to 50%, the degree of stenosis observed was 78% ± 22% for gadolinium-enhanced MR angiography as compared with 69% ± 14% for catheter angiography (p = 0.003).
 View larger version (214K) | Fig. 1A. —76-year-old man with severe hypertension. Radiologic investigation revealed discrepancy between MR angiography and catheter angiography. Catheter angiogram reveals bilateral renal artery stenosis. Right renal artery stenosis (arrow) was considered meaningful using 50% threshold and insignificant using 70% threshold (measurement of investigator 1: 54%; investigator 2: 64%). Similarly, left renal artery stenosis (arrowhead) was meaningful for 50% threshold and insignificant for 70% threshold (investigator 1: 58%; investigator 2: 54%). |
 View larger version (161K) | Fig. 1B. —76-year-old man with severe hypertension. Radiologic investigation revealed discrepancy between MR angiography and catheter angiography. Maximum-intensity-projection reconstruction obtained from gadolinium-enhanced MR angiography shows bilateral renal artery stenoses. However, degree of stenosis was overestimated as compared with catheter angiography. Right renal artery stenosis (long arrow) was considered meaningful using 70% threshold by second investigator (investigator 1, 63%; investigator 2: 71%) and left renal artery stenosis (short arrow) was considered meaningful using 70% threshold by first investigator (investigator 1: 75%; investigator 2: 61%), thus resulting in false-positive results for gadolinium-enhanced MR angiography. |
 View larger version (90K) | Fig. 1C. —76-year-old man with severe hypertension. Radiologic investigation revealed discrepancy between MR angiography and catheter angiography. Intrarenal Doppler waveform obtained from right kidney 1 hr after captopril administration. Systolic rise is perfectly straight (arrow), indicating a normal finding. Left renal Doppler study had equally normal findings. |
Sensitivity and specificity for detecting renal artery stenosis with thresholds of 50% and 70% are reported in Table 1. The sensitivity of gadolinium-enhanced MR angiography was significantly higher than that of captopril-enhanced renal scintigraphy for both the 50% and 70% thresholds both in the kidney and in the patient populations (Tables 2 and 3). The sensitivity of gadolinium-enhanced MR angiography was superior to that of captopril-enhanced Doppler sonography, but the difference reached statistical significance only for the 50% renal artery stenosis threshold. The sensitivity of captopril-enhanced Doppler sonography was significantly greater than that of captopril-enhanced scintigraphy for both 50% and 70% thresholds for renal artery stenosis. No significant difference of the specificity was observed among the modalities, except for gadolinium-enhanced MR angiography (79.5%) and captopril-enhanced Doppler sonography (94.9%), for a 50% threshold for renal artery stenosis in the kidney population (p = 0.03).
Because of a high prevalence of renal artery stenosis in our study population, positive and negative predictive values were estimated using Bayesian analysis for a 30% renal artery stenosis prevalence, as reported in highly selected populations [1]. For detecting renal artery stenosis greater than 50%, positive and negative predictive values were respectively 62% and 86% for captopril-enhanced Doppler sonography, 49% and 76% for captopril-enhanced scintigraphy, and 53% and 98% for gadolinium-enhanced MR angiography. In a nonselected population with a 5% prevalence of renal artery stenosis, positive predictive values were 16%, 10%, and 14% for captopril-enhanced Doppler sonography, captopril-enhanced scintigraphy, and gadolinium-enhanced MR angiography, respectively; and negative predictive values were 98%, 96%, and 100%, respectively.
Three patients had stenosis greater than 50% caused by fibromuscular dysplasia. In this group of patients, renal artery stenosis was accurately diagnosed using scintigraphy in one patient, using Doppler sonography in one patient, and using MR angiography in two patients.
| Discussion | |
|---|
Despite the importance of identifying renovascular disease, a reliable noninvasive technique to detect renal artery stenosis in patients with hypertension has not been clearly established. A reliable noninvasive technique is of paramount importance to select patients with suspected renal artery stenosis for subsequent revascularization procedures and also to avoid unnecessary catheter angiography in patients without substantial renal artery stenosis. Inconsistencies are apparent in the literature regarding noninvasive techniques, especially sonography [16, 17, 19, 31] and radionuclide studies [7, 20, 22]. Consequently, we designed our study to prospectively compare the three leading noninvasive techniques. Most studies define renal artery stenosis as a reduction in diameter greater than 50% based on morphologic evaluation of the renal artery [32]. However, some authors argue that only stenoses greater than 70% should be considered hemodynamically significant [32,33,34]. For this reason, we evaluated the accuracy of each technique using two renal artery stenosis thresholds, 50% and 70%.
Our results show that gadolinium-enhanced MR angiography has a high sensitivity for detection of renal artery stenosis and is probably the most useful noninvasive tool in a high-prevalence population. We observed that normal results on gadolinium-enhanced MR angiography could convincingly exclude renal artery stenosis in 98% of cases. However, gadolinium-enhanced MR angiography has a low positive predictive value (53%) even in a selected population. This low value could be explained by overestimation of the degree of stenosis with gadolinium-enhanced MR angiography as compared with catheter angiography. In fact, the positive predictive value of captopril-enhanced Doppler sonography was greater (62%) than that of gadolinium-enhanced MR angiography. However, we could not reproduce the diagnostic performance of intrarenal captopril-enhanced Doppler sonography reported in earlier studies [13, 15]; this discrepancy may be related to their design, which consisted of a retrospective review of patients who underwent captopril-enhanced Doppler sonography and catheter angiography, without systematic validation with catheter angiography of negative Doppler sonography results. The sensitivity of captopril-enhanced Doppler sonography in our study (analysis by patients) was fair (69%) for detecting renal artery stenosis measuring 50% or greater and good (87.5%) for detecting 70% renal artery stenosis. One limitation of our study is the absence of evaluation of the main renal artery with color Doppler sonography. The accuracy of captopril-enhanced Doppler sonography could be increased by systematically analyzing proximal renal arteries using a direct approach, but this technique is limited by technical failure in 25-42% of cases [16, 17, 35]. Furthermore, this technique is often inadequate for identifying accessory renal arteries, which are present in approximately 20% of patients [16, 17]. On a positive note, there is hope that the use of sonographic contrast agents will improve the evaluation of proximal renal arteries [36, 37].
We observed a poor accuracy for captopril-enhanced scintigraphy for detecting moderate and severe renal artery stenosis, with significantly lower sensitivity than that of gadolinium-enhanced MR angiography and captopril-enhanced Doppler sonography. Our results suggest that captopril-enhanced scintigraphy may not be a useful screening test for renal artery stenosis in populations comparable to ours. In the literature, the diagnostic performance of scintigraphy is much higher than that we report, with results of 51-96% [8]. We have no clear explanation for the poor performance of scintigraphy in our study. All patients had unenhanced and captopril-enhanced examinations performed after the discontinuance of angiotensin-converting enzyme inhibitors, and only 10 patients (25%) had bilateral stenoses. The investigators who performed these examinations are specialists in renal scintigraphy. A possible explanation for our discrepant findings is that most of the results reported in the literature are based on retrospective studies. Those results can be influenced by a verification bias often found in retrospective studies (only patients with positive findings undergo a confirmatory examination). Consequently, the proportion of false-negative examinations may not be evaluated properly, which could lead to an overestimated sensitivity. In our study, the performance of scintigraphy was evaluated prospectively and appears disappointing.
A criticism of our study is the 75% prevalence of renal artery stenosis in our patient population. The admitted prevalence of renal artery stenosis after clinical selection of hypertensive patients is 30% [1, 38]. Therefore, our study population may not be representative of the target population, who should meet the selection criteria that we used. The prevalence of renal artery stenosis among all patients referred to our institution with clinically suspected renal artery stenosis during the study period was 41%. However, some clinicians and patients were reluctant to pursue further examinations in the study protocol after normal findings on Doppler sonography or scintigraphy. In fact, 36 patients in that situation refused to undergo angiography, which explains in large part why the prevalence of renal artery stenosis was further increased in our study population. Another point of debate is that patients examined for renal artery stenosis in our study were selected on suggestive clinical features, which differs from the nonselected population of hypertensive patients, of whom only 2-5% have renovascular disease [30, 38]. Considering this prevalence difference we estimated the predictive values of each screening test in reference to a prevalence as great as 30% (clinically selected patients) and as low as 5% (nonselected patients).
Given the lower availability and higher cost of gadolinium-enhanced MR angiography as compared with captopril-enhanced Doppler sonography, the relative place of these two screening tests remains to be determined. Considering the low positive predictive value of gadolinium-enhanced MR angiography for identifying renal artery stenosis in nonselected patients, the use of this method should be reserved for selected patients. Whether captopril-enhanced Doppler sonography should play a role in the selection remains to be clarified.
In summary, we believe that none of the evaluated modalities could accurately detect renal artery stenosis in nonselected populations. Our study shows that gadolinium-enhanced MR angiography is more accurate than captopril-enhanced Doppler sonography, which is more accurate than captopril-enhanced scintigraphy for the detection of renal artery stenosis. Gadolinium-enhanced MR angiography is the most reliable noninvasive screening test for detecting renal artery stenosis in a population selected on the basis of clinical criteria. The combination of captopril-enhanced Doppler sonography and gadolinium-enhanced MR angiography for screening patients with suspected renovascular disease needs to be evaluated with a cost-effectiveness study.
Supported by operating grant MA15225 of the Medical Research Council of Canada. S. Qanadli was supported by a grant of the Société Française de Radiologie.
Address correspondence to G. Soulez.
| References | |
|---|
