| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ultrasonography |
1 From the University Research Unit B-1068 (Physiopathologic and therapeutic mechanisms of cardiac insufficiency) and the Department of Radiology, Hôpital de Brabois, 54511 Vandoeuvre les Nancy, France (M.C.); the Department of Arterial Hypertension, Hôpital Broussais, Paris, France (P.F.P.); the Department of Radiology, West Glasgow Hospitals University National Health Service Trust, UK (G.M.B.); and Schering, Lys les Lannoz, France (T.R., D.M.D.). From the 1997 RSNA scientific assembly. Received January 15, 1999; revision requested March 1; final revision received May 25; accepted July 23. Supported by a grant from Schering, Berlin, Germany. Address reprint requests to M.C. (e-mail: Michel.Claudon@wanadoo.fr).
 |
Abstract |
|---|
 TOP Abstract IntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: A randomized crossover study was performed in 198 patients from 14 European centers who were referred for renal arterial angiography because they were suspected of having renal arterial stenosis. All patients underwent nonenhanced and SH U 508A–enhanced Doppler US of the renal arteries. Doppler criteria included measurement of renal arterial peak systolic velocity (threshold, 1.4–2.0 m/sec) in all centers and renoaortic ratio (threshold, 3.0–3.5) in nine.
RESULTS: The number of examinations with successful results increased following enhanced Doppler US examination—160 (83.8%) compared with 122 (63.9%) with nonenhanced Doppler US (P = .001), including patients with obesity or renal dysfunction. Renal arterial stenosis (
50%) was detected at angiography in 72 patients. Results at enhanced Doppler US were in agreement with results at angiography more often than with results at nonenhanced Doppler US in the diagnosis or exclusion of renal arterial stenosis (P = .001). For patients examined with nonenhanced and enhanced Doppler US, sensitivity (80.0% and 83.7%, respectively) and specificity (80.8% and 83.6%, respectively) remained unchanged. There were no clinically important adverse events following use of SH U 508A.
CONCLUSION: In patients suspected of having renal arterial stenosis, SH U 508A increased the number of diagnostic renal arterial Doppler studies.
Index terms: Hypertension, renovascular, 961.721 • Renal arteries, stenosis or obstruction, 961.72 • Renal arteries, US, 961.12983, 961.12988 • Ultrasound (US), contrast media, 961.12988 • Ultrasound (US), Doppler studies, 961.12983, 961.12988
|
Introduction |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Although the definitive diagnosis of renal arterial stenosis to date is performed by means of intraarterial digital subtraction angiography, considerable efforts have been made to develop a noninvasive or minimally invasive alternative with sufficient sensitivity and specificity to limit the need for angiography because of its well-documented drawbacks. Intravenous digital subtraction angiography, helical computed tomographic (CT) angiography, captopril renography, and magnetic resonance (MR) angiography have all been assessed in the diagnosis of renal arterial stenosis. Although these techniques have been reported to have sensitivity and specificity exceeding 90% in selected cases, they all have limitations (8,10–15).
Duplex ultrasonography (US) is safe and widely available, but it is time-consuming and operator dependent. One method of diagnosis is to study the intrarenal waveforms (ie, the "tardus et parvus" phenomenon) (16–18). Although this technique is easy to perform, its accuracy is questionable because the lack of an early systolic peak has a low sensitivity for moderate stenoses, and the waveform is dependent on the maintenance of vessel compliance, which limits its effectiveness in elderly patients and patients with atherosclerosis (19,20). Direct Doppler criteria have been proposed for the detection of renal arterial stenosis, including an increased peak systolic velocity and end diastolic velocity at the level of the stenosis, an increase in renal aortic ratio, and the presence of turbulence within the renal artery (21–28). Nevertheless, even in experienced hands, the sensitivity and specificity of direct Doppler analysis of the renal artery is less and the technical failure rate is higher than desirable. Some of the reasons for this include obesity; overlying bowel gas; respiratory motion; depth, tortuosity, and number of renal arteries; and the presence of nephropathy (9,21–23,26).
The use of microbubble echo-enhancing agents has been proposed for increasing Doppler signal intensity in multiple vascular applications. As a galactose-based, air-filled, echo-enhancing agent, SH U 508A (Levovist; Schering, Berlin, Germany), has been shown to increase the Doppler signal level by 20 dB and has proved to be effective in various indications (29–31). Thus, it would be expected to facilitate the visualization of renal arteries, especially in difficult cases.
The purpose of our study was to conduct a multicenter, open-label, randomized crossover trial in patients suspected of having renal arterial stenosis, to evaluate color Doppler flow US examination of the renal arteries prior to and following the intravenous administration of SH U 508A. These results were compared with those from conventional angiography, which was regarded as the standard of reference.
|
MATERIALS AND METHODS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Exclusion criteria were as follows: patients with a renal transplant; patients who had a renal arterial stent and those referred because they were suspected of having renal arterial restenosis; patients who received any iodinated, MR imaging, or US contrast agent in the previous 24 hours or planned to receive these agents within 2 hours of treatment with SH U 508A; lactating women or women known or suspected to be pregnant; patients with galactosemia; and patients with acute myocardial infarction, stroke, severe heart insufficiency, or suspicion of intracranial bleeding.
The study received ethics committee approval according to local legal requirements and was conducted in accordance with the Declaration of Helsinki and European Good Clinical Practice. Written informed consent was obtained from all patients.
Patients were assigned randomly in equal proportions in a crossover fashion to one of two groups. Patients in group 1 (n = 98) underwent nonenhanced Doppler US followed by SH U 508A–enhanced Doppler US. Patients in group 2 (n = 100) underwent enhanced Doppler US followed by nonenhanced Doppler US. To avoid an order effect, the examinations in both groups were performed by two different examiners. The second examiner was blinded to the results of the first examination. All patients underwent angiography—conventional angiography or digital subtraction angiography. Body mass index and creatinine clearance were calculated for each patient.
A range of US machines was used, depending on the center. These included a 128 X P10 (n = 5) (Acuson, Mountain View, Calif); Ultramark 9 HDI (n = 4) and HDI 3000 (n = 1) (Advanced Technology Laboratories, Bothell, Wash); SSA 340A (n = 1) and 140A (n = 1) (Toshiba, Tokyo, Japan); VST Masters (n = 1) (Diasonics, Tirat Carmel, Israel); and AU 590 (n = 1) (Esaote, Genoa, Italy). The same machine, probe, and frequency were used for each patient. The frequency range for the transducers used with the different US machines was 2.5–5.0 MHz. Both color Doppler flow and spectral Doppler examinations were performed.
The general procedure for adjustment of the settings was as follows: (a) Before contrast agent injection, set the filter at its lowest value, set the pulse repetition frequency at a mean value according to the expected normal flow velocity in the investigated vessel without creating aliasing, and adjust the gain to maintain noise level as low as possible. (b) After injection, adjust the gain according to the observed gain intensity increase for optimal filling of the vessel lumen in color mode and delineation of the spectrum envelope in duplex mode, the other settings being kept constant.
Three concentrations of SH U 508A were available: 200, 300, and 400 mg/mL. Preparations were obtained from vials of 2.5 g of substance. On the basis of the results of a brief nonenhanced color Doppler US examination, patients with absent or very low flow signal, defined as demonstration of less than 10% of the artery, received 5 mL of 400 mg/mL; patients with more than very low signal but demonstration of less than 50% of the artery received 7 mL of 300 mg/mL; and those in whom more than 50% but less than 100% of the artery was demonstrated received 10 mL of 200 mg/mL. SH U 508A was administered as an intravenous bolus. After each injection, the catheter was flushed with saline solution. Each patient was allowed to receive up to a maximum of six injections, with a 5-minute minimum delay between each injection. In group 2, there was a gap of 10 minutes between examinations to ensure sufficient washout of SH U 508A.
Methods
The following spectral Doppler diagnostic criteria for renal arterial stenosis were used: (a) localized abnormalities within a portion of the main artery, including increased peak systolic velocity, and presence of turbulence at color Doppler flow (intraluminal alterations of color encoding) or duplex mode (spectral broadening); and (b) increased renal aortic ratio. Owing to differences in stenosis classification between centers, each center was free to use one or several of the above diagnostic criteria and to define pathologic thresholds. Peak systolic velocity with the following thresholds was used in all centers: 1.4 m/sec at one center in 16 patients, 1.5 m/sec at eight centers in 104 patients, 1.8 m/sec at three centers in 54 patients, and 2.0 m/sec at two centers in 24 patients. Renal aortic ratio with the following thresholds was used in nine centers: 3.0 at two centers in 11 patients and 3.5 at seven centers in 105 patients. Each examiner had to report the values of the indexes measured. This allowed a receiver operating characteristic analysis for peak systolic velocity and renal aortic ratio.
Feasibility (ie, the ability to visualize the entire artery to make the US diagnosis or exclusion of renal arterial stenosis) was the primary end point in this study and was analyzed according to both patient and kidney. Investigators were asked to confirm whether it was possible to detect or exclude stenosis greater than 50% in each renal artery. In terms of patients, the analysis was performed as follows: If the presence or absence of stenosis could be evaluated in both the left and right renal arteries, the examination was considered a success; if not, the examination was considered a failure. Thrombosis of the renal artery was considered 100% stenosis.
Secondary efficacy variables included the number of SH U 508A injections required to establish a diagnosis, the duration of each Doppler examination, Doppler signal intensity prior to and following the administration of SH U 508A, the concentration of SH U 508A and its relation to the change in Doppler signal, and the detection of supernumerary arteries. The duration of each Doppler examination was measured from the starting point when the examiner handled the probe to the time at which the Doppler examination of the left and right renal arteries was thought to be completed by the examiner.
Doppler signal intensity prior to and following the administration of SH U 508A was quantified by using the following scales: (a) prior to injection, 0, no usable Doppler signals; 1, low Doppler signal intensity, poor signal-to-noise ratio; 2, moderate but not sufficient Doppler signal intensity; 3, sufficient Doppler signal intensity; (b) after injection, 0, Doppler signal unchanged; 1, slight increase in Doppler signal intensity; 2, considerable increase in Doppler signal intensity; 3, excessive increase in Doppler signal intensity with undesired effects, in spite of appropriate gain setting.
The results of SH U 508A–enhanced and nonenhanced US examinations were compared with those from intraarterial angiography. A hemodynamically significant stenosis was defined as diameter reduction of 50% or more at angiography. The radiologist interpreting the study was blinded to the Doppler examination results.
All statistical tests were two-tailed and performed at the .05 level of significance. The CIs were two-sided with 95% intervals. Efficacy and safety were evaluated from the data obtained in all patients who had at least one injection of SH U 508A. A statistical package (SAS 6-08; SAS Institute, Cary, NC) was used for all calculations. The presence of a center effect was verified by using a logistic regression test (32,33).
Patients were followed up before, during, and 1 hour and 1 day after injection; follow-up included a questionnaire and a physical examination. The onset, duration, intensity, and seriousness of all adverse events—illness, symptom, or unfavorable change in clinical status that appeared or worsened after the start of the study and for 24 hours after the last injection of SH U 508A—were recorded, whether related to the investigational procedures or not. Events were graded as mild, moderate, or severe. The potential relationship to SH U 508A was established according to a five-point scale: not related, unlikely related, possibly related, probably related, definitely related.
|
RESULTS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
|
|
|
|
|
|
Accuracy
Of the 191 patients who underwent angiography following the US examination, 185 underwent digital subtraction angiography and six underwent conventional angiography. Renal arterial stenosis of 50% or greater was detected at angiography in 72 (37.6%) patients, 13 of whom had bilateral stenoses. Renal arterial stenosis was excluded in 119 patients. A renal arterial thrombosis was found in five kidneys. The prevalence of renal arterial stenosis in 382 kidneys at angiographic examination was 17%–43%, except for three centers for which the rate was different (7%, 50%, and 57%).
The agreement between findings at nonenhanced Doppler US, enhanced Doppler US, and angiography for all cases (patients, n = 191; kidneys, n = 382) is presented in Table 3. Regarding patients, enhanced Doppler US was in agreement with angiography significantly more often than it was with nonenhanced Doppler US (134 [70.2%] of 191 patients for enhanced Doppler US vs 100 [52.3%] of 191 patients for nonenhanced Doppler US; P = .001) for the detection or exclusion of renal arterial stenosis. When compared by kidney, the overall results were better and were improved by injection of SH U 508A (299 [78.3%] of 382 kidneys for enhanced Doppler US vs 251 [65.7%] of 382 kidneys for nonenhanced Doppler US; P = .001).
|
|
|
|
|
|
|
Frequency of specific Doppler diagnostic criteria for the evaluation of renal arteries is presented in Table 10: peak systolic velocity measurement was used more frequently (602 [95.0%] of 634 kidneys) than renal aortic ratio (393 [62.0%] of 634); localized flow disturbance was reported in only 31 (4.9%) of 634 kidneys. There was no significant difference in peak systolic velocity, end diastolic velocity, and renal aortic ratio before or after SH U 508A injection, in spite of different examiners performing each examination (P = .461, P = .367, P = .060). There was no difference between the right and the left sides (peak systolic velocity, P = .121; renal aortic ratio, P = .380). The combination of data from both sides was used to calculate sensitivity and specificity.
|
|
Safety
Of the 19 adverse events in 14 patients, seven were estimated to be potentially related to the injection of SH U 508A. They included four cases of the sensation of coldness; one case of paresthesia, or "pins and needles in the hand"; one case of pain at the injection site; and one case of vomiting. There was no severe adverse event.
|
DISCUSSION |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
To our knowledge, the present study is the first randomized multicenter study in a large group of patients (nearly 200) in which renal arterial color Doppler flow US with and without a US contrast agent (SH U 508A) was compared against the reference standard of angiography. The intravenous administration of SH U 508A increased by 20% the number of patients in whom renal arteries could be evaluated (P = .001), including difficult cases such as those involving patients who are obese (body mass index > 27 kg/m2; P = .001) and patients with impaired renal function (creatinine clearance < 50 mL/min [0.83 mL/sec]; P = .033).
Enhanced Doppler US was in agreement significantly more often with angiography than it was with nonenhanced Doppler US, especially for the exclusion of a stenosis. This is an important result if Doppler US is to be considered as a potential screening method; however, this was mainly related to the decrease in the number of nonfeasible examinations.
This improvement in feasibility was achieved without adversely affecting either sensitivity or specificity: sensitivity was 80.0%–83.7%, whereas specificity moderately increased from 80.8% to 83.6% or 86.2%, depending on the subgroups of comparable patients. These results are somewhat lower than those of recently published single-center studies in which the value of Doppler US examination after intravenous injection of contrast agents for the diagnosis of renal arterial stenosis was also evaluated: SH U 508A was used in two (34,35), and a phase-shift agent was used in one (36). Direct diagnostic Doppler signs were used in the evaluation of renal arterial stenosis in two (35,36). Both demonstrated an improvement in sensitivity, which increased from 83% to 95% in one study and from 75% to 100% in the other. However, both of these studies were based on a limited number of patients with a very low feasibility rate of 11% and 12% at baseline examination, respectively. This clearly makes the accuracy of the nonenhanced examinations questionable. In addition, Melany et al (36) reported that, as shown in our study, contrast agent injection did not improve specificity.
This multicenter study involved investigators who were likely to have different levels of experience and used different US machines. After SH U 508A injection, the range of feasibility rates observed in the different centers decreased, with a significant center effect noted in only two centers instead of 10 at baseline.
Supernumerary renal arteries are present in 20%-27% of the population (25, 26). Their detection is of importance, because they can be stenosed either in isolation or in addition to the main artery. However, the prevalence of stenoses occurring in accessory arteries is still unknown, and the physiopathologic consequences that are partly related to the diameter of the stenosed vessel, the extension of the ischemic area, and the presence of functional anastomoses within the kidney are also incompletely understood (25,27). We have shown that the injection of SH U 508A increased the percentage of accessory arteries detected, although the detection rate of 40.8% was lower than that of the 77% reported by Melany et al (36).
SH U 508A was well tolerated and did not compromise the safety of US. This excellent patient tolerance has already been demonstrated in over 2,000 patients worldwide (31). Our experience indicated that, in the case of difficult baseline examination, one injection of contrast agent per side is sufficient in most cases. The lowest concentration, 200 mg/mL, should be used, except in patients with high body mass indexes because they require concentrations of 300 or 400 mg/mL.
There is a debate as to whether contrast agent injection potentially reduces examination duration. In our study, we found a longer mean examination time for enhanced than for nonenhanced Doppler US—25 versus 15 minutes—but this result can at least in part be explained by the design of the study. Lees (35) reported that the use of SH U 508A dramatically reduced the mean examination time from 24.5 minutes for nonenhanced Doppler US to 13.5 minutes for enhanced Doppler US. This advantage could be of potential economic interest if confirmed by subsequent studies.
There are several limitations to the present study, including the lack of strict Doppler criteria for the diagnosis of renal arterial stenosis, as a consequence of the multicenter nature of the trial, and the choice of a percentage of stenosis required to determine a hemodynamically significant lesion.
Doppler criteria may have differed between centers, but the same criteria and corresponding thresholds were used in each individual patient with SH U 508A injection and without SH U 508A injection. The present crossover trial was designed to evaluate the effects of color and duplex Doppler feasibility and diagnosis yield within a given set of diagnostic methods and criteria and not to compare the various Doppler methods and criteria proposed for the diagnosis of renal arterial stenosis. Consequently, the present results can be applied to a large number of currently used US procedures in patients at risk of renal arterial stenosis.
On the other hand, an analysis of previously published single-center studies (21–27) based on direct Doppler evaluation of the renal artery clearly shows that the diagnostic criteria and threshold values fluctuate from one center to the other. Olin et al (25) reported a sensitivity of 98% and specificity of 98% for a peak systolic velocity of 2 m/sec or greater and a renal aortic ratio of 3.5 or greater. With similar criteria, (peak systolic velocity > 1.98 m/sec and renal aortic ratio > 3.3), Miralles et al (27) reported a lower sensitivity of 87.3% and a lower specificity of 91.5%, whereas Helenon et al (26) quoted a sensitivity of 89% and a specificity of 99% with use of a lower peak systolic velocity cutoff point of 1.5 m/sec and taking into account the presence of poststenotic turbulence but not renal aortic ratio. Moreover, a recent intravascular Doppler study (28) showed a sensitivity of 88% and a specificity of 93% with use of a peak systolic velocity threshold of 1 m/sec and a sensitivity of 94% and a specificity of 93% with use of a renal aortic ratio of 1.2. In the current study, we found that renal aortic ratio was more accurate than peak systolic velocity in the diagnosis of a renal arterial stenosis greater than 50%, but it was difficult to determine a precise cutoff point.
In the present study, a stenosis was defined as a luminal reduction in diameter of 50% or more of the renal artery at angiography. A threshold of 50% was chosen—versus a 60% one—because it has been widely used in the recent literature (20,26,32–34,37–39) and was considered hemodynamically significant by most investigators when the study was designed. The accuracy of enhanced Doppler examination could change with use of a 60% threshold to determine a hemodynamically significant stenosis; however, improvement of feasibility by 20% after SH U 508A injection was obtained, regardless of the degree of stenosis in our study.
In conclusion, current recommendations acknowledge that the definitive diagnosis of renovascular disease requires renal angiography, which is costly and carries some risk, particularly contrast agent–induced renal failure or atheroembolism in older patients. Consequently, patients with a clinical suspicion of renal arterial stenosis should undergo noninvasive tests to avoid unnecessary angiography, if those tests can be used to identify renal arterial stenosis accurately (40). In the present trial, we found that SH U 508A safely and significantly converted 20% of nondiagnostic color Doppler flow examinations into diagnostic examinations. Since baseline and SH U 508A–enhanced Doppler examinations yielded similar accuracy, SH U 508A-enhanced Doppler examination reduced the number of inconclusive examinations by 20% and might reduce in a similar proportion the number of unnecessary angiograms. This is of major interest in patients with impaired renal function for whom SH U 508A injection has proved to increase the feasibility rate significantly.
The review of the literature in conjunction with the conclusions of the current multicenter study demonstrate that there is a need for establishing a consensus opinion regarding Doppler signs, useful criteria, and thresholds for the diagnosis of renal arterial stenosis in a manner analogous to that of the North American Symptomatic Carotid Endarterectomy Trial for internal carotid arterial stenosis (41). This consensus has to be reliable, regardless of the US equipment used.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Author contributions: Guarantors of integrity of entire study, M.C., T.R., D.M.D.; study concepts, T.R.; study design, M.C., T.R.; definition of intellectual content, M.C., P.F.P.; literature research, M.C.; clinical studies, M.C., G.M.B.; data acquisition, T.R.; data analysis, T.R., M.C., P.F.P.; statistical analysis, D.M.D.; manuscript preparation, M.C., T.R., D.M.D.; manuscript editing, M.C., D.M.D., G.M.B.; manuscript review, P.F.P., G.M.B.
|
References |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  G. J. Williams, P. Macaskill, S. F. Chan, T. E. Karplus, W. Yung, E. M. Hodson, and J. C. Craig Comparative Accuracy of Renal Duplex Sonographic Parameters in the Diagnosis of Renal Artery Stenosis: Paired and Unpaired Analysis Am. J. Roentgenol., March 1, 2007; 188(3): 798 - 811. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. E. Zierler Atherosclerotic Renovascular Disease: Natural History and Diagnosis Perspectives in Vascular Surgery and Endovascular Therapy, December 1, 2004; 16(4): 299 - 310. [Abstract] [PDF]
|
|
|
|
 |
|
 G. B. C. Vasbinder, P. J. Nelemans, A. G.H. Kessels, A. A. Kroon, P. W. de Leeuw, and J. M.A. van Engelshoven Diagnostic Tests for Renal Artery Stenosis in Patients Suspected of Having Renovascular Hypertension: A Meta-Analysis Ann Intern Med, September 18, 2001; 135(6): 401 - 411. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
) and the entire right renal artery (arrow), which was reported as considerable by examiner B. Aliasing and turbulence are detected in the proximal artery. (c) Right oblique anterior duplex Doppler US image depicts waveforms obtained at proximal right renal arterial level (arrow) that show spectral broadening and increased peak systolic velocities greater than 2.5 m/sec. (d) Subsequent right oblique anterior angiogram shows a severe stenosis (arrow) of the proximal right renal artery.
