| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Institute of Clinical Radiology and Nuclear Medicine, University Medical
Center Mannheim, Medical Faculty of Mannheim, University of Heidelberg,
Theodor-Kutzer-Ufer 1-3, 68167 Mannheim, Germany.
2 Department of Radiology, Catholic Hospital, Koblenz, Germany.
3 Department of Medical Statistics, Biomathematics, and Information Technology,
Medical Faculty of Mannheim, University of Heidelberg, Heidelberg,
Germany.
Received September 11, 2008;
accepted after revision October 14, 2008.
H. J. Michaely is a consultant to Bayer HealthCare. D. J. Dinter and K. W.
Neff contributed equally to this study.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Peripheral bolus-chase MRA runoff studies in 177 patients with peripheral arterial occlusive disease (PAOD) were retrospectively assessed with regard to the rate of nondiagnostic image quality due to substantial venous overlay in the calf arteries requiring repeated MRA examinations. Logistic regression was used to analyze the rate of nondiagnostic MRA examinations as a function of several endogenous and exogenous risk factors and of the stage of PAOD. To probe the retrospective data, 22 consecutive patients were prospectively included and underwent a standard peripheral MRA examination if the probability of a nondiagnostic examination was less than 50% based on the results of logistic regression; otherwise, a hybrid MRA examination was ordered.
RESULTS. Nondiagnostic image quality of the calf arteries was found in 53 patients (30%). The incidence increased with each stage of PAOD up to 39% for stage IV. For each increase in the stage of PAOD, the probability of nondiagnostic image quality increased by a factor of 1.5561 (p = 0.0024). With an increasing number of risk factors, a significantly (p = 0.0074) higher rate of nondiagnostic images was found.
CONCLUSION. Based on the retrospective statistical analysis of PAOD stages and risk factors, selected patients can be triaged to undergo a specific hybrid MRA technique and thus circumvent the occurrence of nondiagnostic images and the need for repeated MRA examinations.
Keywords: MR angiography • peripheral arterial occlusive disease • peripheral vessels • runoff study • second injection • venous overlay
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
A technically adequate MRA examination is the prerequisite for such excellent results. Most commonly, a bolus-chase technique is used. Here, MRA acquisitions are performed consecutively at several anatomic levels from the body toward the periphery [3]. More recently, with the advent of whole-body MR scanners, runoff studies have been integrated into a whole-body MRA examination [8, 9]. Bolus-chase techniques are technically relatively easy to apply and require only a single injection of contrast agent. However, their main limitation is the high probability of substantial venous overlay in the calf station, particularly in patients with advanced stages of PAOD or in diabetic patients. Therefore, several solutions have been presented to overcome this limitation. Venous return can be reduced by placing a pressure cuff around the patient's thighs [10] or by adjusting the scanning parameters to better match the movement of the table with the flow of the contrast agent down the patient's legs [11]. Another solution is to change the MRA acquisition scheme to a so-called "hybrid approach" in which images are acquired at the calf station first after the initial contrast administration. After a second injection of contrast medium, the remaining parts of the peripheral vessels are imaged using a bolus-chase approach [12].
|
Currently, no evidence-based recommendations are available to identify patients who are more likely to require a hybrid technique MRA or those who need standard bolus-chase MRA. Therefore, the aim of this study was to comprehensively analyze the individual risk factors for producing images of nondiagnostic quality due to substantial venous overlay and to prospectively probe the findings of this retrospective analysis.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
For the prospective part of the study, 22 consecutive inpatients (14 men and eight women; mean age, 71.7 ± 9.3 years) were included between July 11, 2007, and March 20, 2008, after providing informed consent. Patients with an estimated GFR of less than 30 mL/min (MDRD study formula) were excluded from the prospective study to prevent the development of NSF. A standard bolus-chase MRA with a single injection of contrast agent, as described earlier, was prescribed when the probability of obtaining nondiagnostic image quality as based on the retrospective data analysis was less than 50%; a hybrid MRA was chosen when the probability for venous overlay was calculated to be higher than 50%. A flowchart of the study is presented in Figure 1.
Determination of Risk Factors and Image Evaluation
For all patients, the presence and the maximum degree of PAOD as determined
by the referring clinician were registered according to the Fontaine lower
limb ischemia classification: stage 0, no PAOD; I, mild claudication and
patient can complete treadmill exercise; II, moderate claudication; III,
severe claudication and patient cannot complete treadmill exercise; and IV,
ischemic rest pain. On the basis of the patients' medical files, the presence
of established cardiovascular risk factors was determined and entered into an
electronic score sheet in a dichotomous way (0, risk factor not present; 1,
risk factor present). Complete medical records were available for 132 of the
177 patients.
The American Heart Association (AHA)-acknowledged causative risk factors for cardiovascular disease—that is, diabetes mellitus, hypertension, hyperlipoproteinemia, and cigarette smoking—were chosen as endogenous risk factors [15]. Other endogenous risk factors included the following concomitant diseases: acute renal failure, chronic renal failure, osteomyelitis, phlegmon of the feet and calves, the presence of arterial bypasses, and the presence of allogenic material such as stents or bypass grafts. Concomitant medications were considered exogenous risk factors and included insulin, antihypertensive drugs, aspirin, statins, and coumarin. The endogenous risk factors chosen were based on AHA recommendations [15], results from previous studies [13, 14], and results from the Prospective Cardiovascular Münster (PROCAM) study [16]. They represent medications commonly administered to patients with PAOD that might alter the hemodynamics. All data were entered into a digital score sheet using Excel (Excel 2003, Microsoft).
For the prospective part of this study, the image quality of the calf station was evaluated by two radiologists (10 and 7 years of experience in vascular MRI) in consensus. The aforementioned Likert scale was used.
MR Protocol and Image Analysis
All MR examinations were performed on an 8-channel 1.5-T MR scanner
(Magnetom Sonata Maestro Class, Siemens Healthcare) equipped with a
high-performance gradient system with a maximum amplitude of 40 mT/m and a
slew rate of 200 mT/m/ms using a dedicated circularly polarized peripheral
coil.
According to the protocol, localizing sequences and mask images were acquired before the contrast agent was administered. An automated injector pump (Spectris, Medrad) was used to inject contrast agent at a dose of 0.2 mmol/kg of body weight. Gadopentetate dimeglumine (Magnevist, Bayer Schering Pharma) was injected at a flow rate of 2 mL/s and was followed by a 30-mL saline chaser at the same flow rate through a standard 18-gauge needle placed in an antecubital vein. A fluoroscopic triggering technique in the coronal direction was applied to detect the arrival of the contrast agent bolus at the level of the renal arteries. Immediately after the contrast medium was visible at this level, three-station bolus-chase MRA of the lower extremity was manually started. For all three stations, fast 3D spoiled gradient-echo sequences were used. The exact sequence parameters are given in Table 1. All MRA examinations were supervised by an attending physician with 10 years of experience in vascular imaging.
|
After the MRA acquisition, the quality of the MRA examination in the calf station was assessed using a 4-point Likert-type scale, with 1 representing nondiagnostic (complete venous overlay precluding diagnostic reading); 2, moderate (moderate venous overlay hampering but not precluding diagnostic reading); 3, good (minimal venous enhancement not interfering with diagnostic reading); and 4, very good (no venous enhancement). When the image quality was nondiagnostic at the calf station—that is, a Likert score of 1—the attending radiologist ordered a repeated focused MRA acquisition of the calf station. A repeated MRA examination was also ordered if images of only one side were nondiagnostic. This MRA examination was repeated during the same session at least 10 minutes after the initial contrast injection using the same sequence.
For the second MRA examination, a bolus of gadopentetate dimeglumine was administered at a dose of 0.1 mmol/kg of body weight. Fluoroscopic triggering was performed at the level of the popliteal artery to ensure that bolus arrival was correctly synchronized with the acquisition of k-space center. The total contrast agent dose for the repeated MRA examination together with the initial MRA examination was 0.3 mmol/kg of body weight of gadopentetate dimeglumine.
For the prospective part of the study, the same sequences with parameters identical to those in the retrospective study were used. After determining the number of endogenous and exogenous risk factors, an attending physician (10 or 7 years of experience in vascular imaging) decided if a standard bolus-chase MRA examination or if a hybrid MRA examination of the lower extremities should be prescribed.
A standard bolus-chase MRA with a single injection of contrast agent, as described earlier, was performed when the probability of obtaining nondiagnostic image quality due to substantial venous overlay based on the retrospective data analysis was less than 50%; otherwise, hybrid MRA was chosen. The hybrid MRA images at the calf station were acquired first after administering the bolus of gadopentetate dimeglumine (0.1 mmol/kg of body weight) with a flow rate of 2 mL/s followed by a 30-mL saline chaser. Fluoroscopic triggering was performed at the level of the popliteal artery to ensure that bolus arrival was correctly synchronized with the acquisition of k-space center. Images of the pelvic and thigh stations were then acquired as described earlier, resulting in a total dose of 0.3 mmol/kg of body weight of gadopentetate dimeglumine.
Statistical Analysis
All statistical analyses were performed using SPSS (version 13.0, SPSS) and
SAS (release 9.01, SAS Institute) software. The incidence of repeated MRA
acquisition was determined with regard to the presence of various risk factors
and the stage of PAOD. To compare the two groups regarding different risk
factors, the chi-square test was used. In addition, odds ratios for a required
second contrast agent injection were calculated separately based on the stage
of PAOD and the number of risk factors. Furthermore, as a multiple statistical
model, logistic regression with stepwise selection was applied to identify
risk factors and covariates associated with the need for a second injection.
Effects with a p value of less than 0.05 were considered to be
statistically significant. The results of the image analysis for the
prospective study are presented as medians because of their ordinal
nature.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
Influence of Endogenous Risk Factors
Hypertension was the only single risk factor for which a statistically
significant increase was found in the rate of nondiagnostic MRA examinations
of the calf arteries. In patients without hypertension, nondiagnostic MRA
examination of the calf arteries was present in 11% (3/27 patients), whereas
among those with hypertension, image quality was nondiagnostic in 32% (34/105
patients) (p = 0.0282). A detailed overview of all risk factors is
shown in Table 3.
|
Influence of Exogenous Risk Factors
There were no major differences among the different types of medications
being analyzed in this study and the number of nondiagnostic MRA examinations
of the calf arteries. With all drugs studied, the rate of nondiagnostic MRA
examinations of the calf arteries was between 22% (coumarin) and 30%
(insulin). In patients taking statins at the time of the examination, the rate
of repeated MRA examinations was 28%; with aspirin, 29%; and with
antihypertensive drugs, 30%. There was no significant association between the
current medication and the rate of nondiagnostic MRA examinations of the calf
arteries.
Additive Effect of Multiple Risk Factors
An increasing rate of nondiagnostic MRA examinations of the calf arteries
was observed with a rising total number of risk factors. In four patients
without any risk factors, only diagnostic images were obtained. In 37 patients
with one risk factor, the image quality of the calf arteries was nondiagnostic
for five MRA examinations (14%), whereas in the 58 patients with two risk
factors this occurred in 19 patients (33%). In patients with three risk
factors (n = 33), the rate of nondiagnostic MRA examinations of the
calf arteries reached 39%. The trend toward a higher rate of nondiagnostic MRA
examinations of the calf arteries in patients with a higher number of risk
factors was statistically significant (p = 0.0074, logistic
regression). The odds ratio was assessed as 2.043, indicating that the
relative risk for nondiagnostic image quality of the calf arteries due to
venous overlay roughly doubles with each additional risk factor.
Multiple Logistic Regression Analysis
Based on the retrospectively analyzed patient data and by means of a
logistic regression model, the following formula for the probability of
repeated MRA was obtained:
 |
|
Venous overlay preventing diagnostic image reading was not found. The median image quality of the calf station was 3—that is, good (minimal venous enhancement not interfering with diagnostic reading—in patients who underwent a standard bolus-chase MRA examination. In patients assigned to a hybrid MRA examination, the median image quality of the calf station was 4—that is, very good (no venous enhancement).
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Our results show that the rate of repeated examinations increased steadily with the advance in PAOD stage. Based on our retrospective data, the relative risk for a nondiagnostic MRA examination of the calf arteries in a patient with PAOD stage IV is increased by a factor of 5.94 compared with the risk for nondiagnostic MRA in a patient without PAOD. Apart from hypertension, which was associated with a significantly higher rate of repeated injections, the other single risk factors—among them, nicotine abuse, osteomyelitis, and diabetes mellitus—did not increase the risk for repeated MR examinations. This result was surprising particularly with regard to diabetic patients and patients with osteomyelitis given that time-resolved imaging approaches have been proposed because venous contamination occurs more frequently in those populations [17]. Another possibility that may explain this finding could be the relatively small number of patients in this study, which also represents a limitation. Therefore, the lacking statistical significance could be attributed to a statistical error of the second kind (i.e., an insufficient number of patients). In practice, most patients will have more than one risk factor.
The analysis of our data proves that the number of nondiagnostic MRA examinations due to venous contamination increased significantly with the number of simultaneous endogenous risk factors in a given patient, whereas the presence of exogenous risk factors did not seem to influence the outcome of the MRA examination. In addition, the results of this study show that multiple risk factors have an additive character in terms of image degradation. Previous studies on this topic addressed the potential risk factors separately [13, 14]. The exponential slope of the logistic regression model shows this additive character of multiple risk factors.
For daily MR examination routine, it would be desirable to be able to anticipate the risk of nondiagnostic MRA examination of the calf arteries before starting the MRA procedure. For this purpose, a predictive model for the probability of venous overlay based on the retrospective data according to the stage of PAOD and the number of simultaneously occurring endogenous risk factors was established. This model can serve as a means of triaging patients to choose the appropriate MRA technique. In patients with no to little risk for venous contamination, a standard bolus-chase technique can be chosen. However, patients with a high risk of venous contamination should undergo a hybrid MRA or a time-resolved MRA study of the lower legs [12, 17-20].
In the prospective part of our study, a probability of 50% was chosen as the cutoff for a high risk of venous contamination, which was the case in 32% of the prospectively included patients. The fact that none of the examinations of these patients revealed venous enhancement can be seen as a proof of concept for the prediction model. In addition, the median image quality score was 4 with the hybrid approach versus 3 with the conventional bolus-chase MRA, which again supports application of the prediction model. Most patients who underwent standard bolus-chase MRA showed slight venous enhancement that did not interfere with diagnostic reading. Some slight venous enhancement might, however, be seen as an inherent characteristic of the bolus-chase technique that, in most cases, does not impair the diagnostic reading. In the intermediate-risk group, the venous return can be decreased by using a blood pressure cuff with subsystolic pressure [10, 21]. This technique has the advantage of being cheap and easily applied. It might, however, decrease patient comfort.
Our approach of selecting patients with a high likelihood for having examinations with venous contamination before MRA examination has the additional advantage of being efficient with regard to contrast agent administration. Both the repeated MRA examination, performed because of nondiagnostic images of the calf arteries, and the hybrid approach require the injection of 0.3 mmol/kg of standard gadolinium chelates. With the a priori knowledge about which patients would benefit from the hybrid MRA, the number of patients allocated to this procedure can be minimized.
MRA is considered a standard technique for the detection and grading of PAOD and is recommended by current international guidelines as the technique of choice because of its noninvasiveness and high accuracy [4]. Only little information is provided about the suggested MRA parameters. A spatial resolution of 1.0 x 1.0 x 1.0 mm should be achieved in the lower leg [4]. No specifications are given for the upper leg or pelvis. In the current study, we used parallel imaging only for the calf station with a spatial resolution of 1.2 x 0.8 x 1.0 m, whereas the spatial resolutions of the thigh and pelvis stations were slightly lower.
By applying parallel imaging with smaller voxel sizes, new coil concepts such as the total imaging matrix and particularly high field strengths will foster the use of MRA as an accurate means of diagnosis [11, 22]. If a lower spatial resolution is acceptable, the MRA acquisition can be sped up at the cost of diagnostic accuracy. For the thigh station, for example, thicker partitions can be chosen, as suggested by Maki et al. [11]. Therefore, the problem of early, disturbing venous return will prevail unless completely different imaging strategies, such as time-resolved MRA, intravascular contrast agents, or both, are chosen. The commercially available intravascular contrast agent gadofosveset trisodium—previously known as MS-325—offers the possibility of acquiring time-resolved MRA data during the first pass and ultra-high-spatial-resolution MRA data during the steady state. In a recent study, investigators reported that an acquired spatial resolution with a 0.5-mm isotropic voxel size [23] provided excellent visualization and differentiation of the arteries and veins of the lower leg.
Our study has some well-recognized limitations. Because of the retrospective design of the study, we were not able to retrieve all of the patients' records, so a complete history of risk factors could be obtained in only 132 of the 177 patients. The model that we suggested for calculating the probability of nondiagnostic MRA examination of the calf arteries is also based on these retrospective data, but it was successfully probed in the prospective part of this study. These findings reflect primarily the patient population at our institution but should be largely applicable for any other hospital of this size. However, these results may not be valid for other institutions with a larger number of outpatients who are less critically ill.
In conclusion, the current study is the first to quantitatively and comprehensively assess the number of nondiagnostic MRA examinations of the calf arteries with regard to the stage of PAOD and to a set of common cardiovascular risk factors. According to our results, patients for whom the probability of a nondiagnostic MRA examination of the calf arteries is high can be preselected and hence patients can be triaged to the appropriate MRA technique.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
