| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Vascular and Interventional Radiology |
1 From the Department of Radiology–MRI, New York University Medical Center, 530 First Ave, HCC Basement, New York, NY 10016. Received August 26, 1998; revision requested October 22; final revision received February 8, 1999; accepted June 8. Supported in part by an RSNA Research and Education Foundation Medical Student Departmental Grant. Address reprint requests to V.S.L. (e-mail: lee@mri.med .nyu.edu).
 |
Abstract |
|---|
 TOP Abstract IntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: A retrospective study included 15 patients who underwent 3D gadolinium-enhanced subtraction MR venography with a spoiled gradient-echo sequence before and at multiple times after intravenous administration of 30-40 mL of contrast material. Maximum intensity projection and multiplanar reconstruction images were used to categorize central veins as patent, occluded, or narrowed. Results were compared with findings (in 12 patients) at conventional venography (n = 3), attempted central venous catheter placement (n = 3), or surgery (n = 6). Medical records were retrospectively reviewed to determine if patient care was affected by MR venographic findings.
RESULTS: By using MR venograms, an appropriate vessel could be identified for successful placement of a catheter, indwelling venous access device, or arteriovenous hemodialysis graft in all nine patients in whom placement was attempted. MR venography also was predictive of unsuccessful hemodialysis catheter placement in one patient. Conventional venographic findings confirmed MR venographic findings in three patients; in a fourth patient, conventional venography was unsuccessful due to inadequate access. MR venographic findings influenced treatment in 14 patients.
CONCLUSION: On the basis of these initial results, 3D gadolinium-enhanced MR venography may facilitate comprehensive evaluation of abnormalities of the central veins in the thorax, particularly with regard to selection of venous access sites.
Index terms: Gadolinium • Magnetic resonance (MR), vascular studies, 94.129412, 94.12942, 94.12943 • Veins, access, 94.1269 • Veins, MR, 94.129412, 94.12942, 94.12943 • Veins, stenosis or obstruction, 94.75 • Veins, thrombosis, 94.75
|
Introduction |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Some researchers (6–13) have advocated magnetic resonance (MR) imaging as an alternative to conventional venography for evaluation of the central veins. Whereas, to date, most investigators (6–12) have used two-dimensional (2D) time-of-flight (TOF) methods, Lebowitz et al (13) recently described a faster approach involving gadolinium-enhanced three-dimensional (3D) gradient-echo MR imaging with subtraction of arterial phase images from delayed phase images to obtain gadolinium-enhanced venograms. This approach can potentially provide a comprehensive evaluation of central veins within a few breath holds. In addition, as with 2D TOF techniques, patients with renal insufficiency can benefit from this method because the contrast agent, gadopentetate dimeglumine, is not nephrotoxic. We sought to evaluate the feasibility and usefulness of this method in the evaluation of patients referred because of possible abnormalities of the central venous system.
|
MATERIALS AND METHODS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Eight patients had chronic renal failure (mean serum creatinine level, 7.7 mg/dL [681 µmol/L]; range, 4.0–11.8 mg/dL [354–1,043 µmol/L]). Seven patients had a documented history of venous thrombosis and/or multiple failed attempts at central venous access. Clinical indications for MR venography were (a) surgical planning for potential creation of an arteriovenous hemodialysis graft (n = 7), (b) symptomatic central venous thrombosis (n = 3, including superior vena cava [SVC] syndrome in two), and (c) evaluation for central venous access or for cardiac intervention (n = 5, including access for placement of a hemodialysis catheter, hyperalimentation catheter, and indwelling venous access device for chemotherapy and preprocedural planning for a cardiac electrophysiologic study and pacemaker placement). Patients dependent on dialysis underwent dialysis within 24 hours of MR imaging.
All patients underwent MR imaging with a 1.5-T unit (Magnetom Vision; Siemens Medical Systems, Iselin NJ) equipped with a torso phased-array coil. Informed consent for administration of the gadolinium-based contrast material was obtained from all patients before imaging, according to standard procedure at our institution. Gadolinium-enhanced venography was performed in all patients in the coronal plane by using a 3D spoiled gradient-echo sequence during suspended respiration.
Sequence parameters evolved during the course of this study as newer and faster sequences became available. The parameters used for the first seven patients were as follows: 5/2 (repetition time msec/echo time msec), 45°–50° flip angle, 375–475-mm rectangular field of view (determined on the basis of patient body habitus), 64–179 x 160–256 matrix, 102–230-mm slab thickness, 2.2–3.0-mm section thickness, mean acquisition time of 29.5 seconds ± 5.7 (SD). For the next eight patients, a modified sequence that provided interpolation in the section-selective direction was used, with the following parameters: 4.2/1.7 and flip angle of 25° (n = 3) or 3.8/1.3 and flip angle of 25°–40° (n = 5), 324–500-mm rectangular field of view (dependent on body habitus), 128–166 x 160–256 matrix, 84–122-mm slab thickness, 1.5–2.5-mm interpolated section thickness, mean acquisition time of 20.6 seconds ± 1.1.
In all patients, 3D gradient-echo images were acquired both before and at multiple times after intravenous administration of 30 mL (n = 1) or 40 mL (n = 14) of gadopentetate dimeglumine (Magnevist; Berlex Laboratories, Wayne, NJ) (mean dose, 0.28 mmol/kg). We routinely administer 40 mL of gadolinium-based contrast material for MR venography, provided the dose does not exceed 0.3 mmol/kg, in which case 30 mL is used. All injections were administered with an MR compatible power injector (Spectris; Medrad, Pittsburgh, Pa) through either a peripheral 22-gauge intravenous catheter or a central venous catheter (patient 13) at a rate of 2 mL/sec, immediately followed by 20 mL of normal saline solution (also at 2 mL/sec).
Before administration of contrast material, a timing examination was performed according to the method of Earls et al (14). Briefly, a 1-mL test dose of gadopentetate dimeglumine followed by 20 mL of normal saline solution was infused at a rate of 2 mL/sec, and axial magnetization-prepared spoiled gradient-echo MR images (7.7/4.2/300 [repetition time msec/echo time msec/inversion time msec], 15° flip angle) were acquired every 2 seconds in a fixed location to estimate patient circulation time (time to peak arterial enhancement). On the basis of the timing examination results, acquisition of the first of at least three contrast-enhanced 3D gradient-echo MR data sets was synchronized to maximal arterial enhancement, with subsequent acquisitions after a 20–30-second delay (Fig 1). All images were acquired during suspended respiration at end expiration.
|
|
|
|
MR imaging results were correlated with findings at conventional venography (n = 3), attempted catheter placement (n = 3), or surgery (n = 6). No confirmatory data were available in three patients. For each correlative study or operation, MR imaging findings could be confirmed only for those veins (and their corresponding deep draining veins) that were injected, had a cannula inserted, or were used for arteriovenous graft placement. Patient medical records were retrospectively reviewed to determine if clinical management was affected by MR venographic results.
|
RESULTS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
|
In the second patient who underwent conventional venography (patient 1), multiple attempts to place a central venous catheter via the right subclavian and right external jugular veins for hyperalimentation had failed prior to diagnostic imaging. Blood return was achieved each time; however, the attempts were aborted because the guide wire could not be advanced smoothly. MR venograms showed patent right subclavian and right brachiocephalic veins with a nonocclusive web at the junction of the two veins. Both right and left internal jugular veins were thrombosed. Conventional venographic findings confirmed patency of the right subclavian and brachiocephalic veins but did not demonstrate the web; the jugular veins were not evaluated. With fluoroscopic guidance, a percutaneous central venous catheter was successfully placed on the right side.
The third patient who underwent conventional venography (patient 4) had a history of congenital heart disease and multiple catheterizations and underwent chest MR venography and separate pelvic MR venography to evaluate access sites for a cardiac electrophysiologic study. The pelvic MR venogram revealed occlusion of the right femoral and iliac veins and patent left femoral and iliac veins with an occluded inferior vena cava just below the level of the renal veins. These findings were corroborated during bilateral femoral vein catheterizations. The chest MR venogram revealed a completely patent thoracic central venous system. The left subclavian vein was later used successfully as the access site for an electrophysiologic study.
In all nine patients in whom arteriovenous graft placement or catheterization was attempted, MR venographic results helped guide successful intervention (Table). After placement, all arteriovenous grafts or catheters were used successfully. MR venographic findings also were predictive of one unsuccessful attempt at hemodialysis catheter placement in a patient in whom peritoneal dialysis was subsequently initiated (patient 11, Fig 2). Patient 10 had a malfunctioning left arteriovenous hemodialysis graft and was considered for construction of a new graft in the right arm. After MR venography demonstrated diseased right axillary, subclavian, internal jugular, and brachiocephalic veins, however, the patient underwent successful repair of the occluded left-sided graft, and subsequent hemodialysis was successful with the repaired graft.
In three patients, no corroboration for the MR venographic results was available. Two of these patients (patients 7 and 15) underwent MR venography to help diagnose central venous occlusion, and one of these two (patient 15) underwent successful anticoagulation therapy. In patient 7, anticoagulation therapy was withheld because of the patient's low platelet count. In the third patient (patient 5), MR venography demonstrated occlusion of nearly all of the thoracic central veins (Table), and a clinical decision was made to place a pacemaker at anterior thoracotomy.
Overall, MR venographic findings influenced medical care in 14 of 15 patients in this study (Table). In no patient was treatment misguided on the basis of MR venographic results.
|
DISCUSSION |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Conventional venography, although considered to be the reference standard for evaluation of the central veins (15), has limitations. These limitations include risks of nephrotoxicity and allergic reaction associated with iodinated contrast material and exposure to ionizing radiation. Conventional venography also is limited because venous access must be established on the same side as the suspected abnormality. Furthermore, multiple access sites must be established when more than a single drainage system must be assessed.
Alternative methods to assess the patency of central veins include ultrasonography (US) and contrast-enhanced CT. US evaluation may be hindered because the pertinent venous structures are obscured by osseous structures and lung parenchyma (16–18). Contrast-enhanced CT can be used to evaluate central veins, but it has disadvantages associated with the use of iodinated contrast material, as well as artifacts that can occur during contrast material administration (19,20). To minimize artifacts and increase accuracy, CT evaluations typically require injections in veins in both arms. In one patient in this study who subsequently underwent diagnostic MR venography (patient 6), limited peripheral venous access resulted in poor venous contrast enhancement and an inconclusive CT scan.
Several groups (6–13) have investigated the use of MR imaging for evaluation of the central venous system. To date, most (6–12) have used a 2D TOF approach, which has the advantage of not requiring contrast material; results from these studies have established an excellent correlation with results from conventional venography. In a study by Finn et al (6) with 30 patients who were suspected of having thoracic venous occlusion, 2D TOF MR venographic results were completely corroborated in 19 (86%) of 22 cases, and MR imaging findings were predictive of successful venous catheterization in 11 patients. Similarly, Hartnell et al (8) reported agreement between 2D TOF MR venographic and conventional venographic findings in all 17 cases where conventional venographic correlation was available. Moreover, successful surgical catheter placement was predicted on the basis of 2D TOF MR venographic findings in all 28 cases in which such intervention was attempted; 2D TOF MR venographic findings also were predictive of two unsuccessful attempts at venous cannula placement.
To minimize in-plane saturation at TOF imaging, acquisitions must be performed orthogonal to the direction of flow. The jugular veins and SVC can be evaluated with a series of axial sections, but the subclavian and axillary veins are better evaluated by using oblique sagittal or axial acquisitions (6–8). If saturation bands are used for selective demonstration of veins, the venous signal intensity in collateral (ie, retrograde) pathways may inadvertently be suppressed. Moreover, venous flow in the arms (when placed at the patient's side) is cephalic and requires a separate acquisition from that used for the SVC and jugular veins, where flow is caudal. To solve this problem, Rose et al (7) had patients abduct their arms above the head for axial acquisitions, possibly at the expense of patient comfort.
Artifacts also may hinder interpretation of 2D TOF MR venograms. Finn et al (6) noted occasional focal signal intensity decreases at venous confluences and in vessels with a long in-plane course; such decreases could be misinterpreted as thrombi. They emphasized the importance of reviewing all individual sections from the source data to accurately assess the patency of central veins. This can be cumbersome, however, and, in a study by Rose et al (7), the large number of images that needed to be reviewed was partly responsible for an interobserver interpretation variability rate of 44%. Finally, the 2D TOF technique also is limited because collapsed veins with little or no flow are not easily demonstrated (21).
In a recent study, Li et al (21) described the use of direct peripheral venous injection of low doses of gadolinium-based contrast material for the detection of deep venous thrombosis in three patients and two volunteers. As compared with the 2D TOF technique, their approach provided sharper images in a shorter imaging time. As with conventional venography, however, venous access in the limb of interest was necessary for administration of contrast material, and a single, unilateral drainage system typically was assessed.
The use of gadolinium-based contrast material bypasses the dependence on inflow and allows imaging with large fields of view in the coronal or sagittal plane, despite substantial in-plane venous flow. This 3D approach can thereby yield a fast and comprehensive evaluation of all veins included in the image field of view. Lebowitz et al (13) provided an early description of the technique of subtraction of selective arterial phase images from subsequently acquired delayed phase images to obtain 3D selective MR venograms. In their preliminary investigation with 17 patients, MR studies in four were abnormal, most commonly revealing deep venous thrombosis of the lower extremities. Our results demonstrate the feasibility of this approach of indirect venography as applied in patients who need evaluation of central veins for access or for hemodialysis graft placement.
Because 3D gadolinium-enhanced MR venography is reliant on recirculation of contrast material in the venous system, peripheral intravenous access can be established anywhere in the body. This approach is, therefore, readily applied in patients with limited intravenous access. This flexibility enabled one patient (patient 14) in our study to undergo successful 3D gadolinium-enhanced MR venography after attempted conventional venography failed. Gadolinium-based contrast material also can be safely used in patients with renal insufficiency and has a substantially lower risk of nephrotoxicity, as compared with that of iodinated contrast material (22,23). Moreover, use of MR venography obviates patient exposure to ionizing radiation.
There are limitations to this approach. Gadolinium-enhanced 3D MR venography requires some form of peripheral venous access, which is not available in all patients. One of the patients in our study (patient 11) was referred for repeat MR venography 3 months after the initial examination and underwent a 2D TOF examination because of a lack of venous access. However, as a consequence of the limitations associated with a 2D TOF approach, as described earlier, evaluation of subclavian venous patency was limited in this patient owing to in-plane saturation effects (Table). Another limitation was that the first pass of contrast material may be associated with signal intensity decreases in and around the veins on the side of injection, due to the T2 effects that predominate at high concentrations. Typically, this primarily affects the arterial phase image. Also, gadolinium-enhanced MR venography yields no information about flow direction (such as in the azygos vein), although this information can be obtained with subsequent selective TOF or phase-contrast MR techniques.
Subtraction techniques offer a selective demonstration of veins, but a vein-free arterial study must be obtained first. Furthermore, precise registration of the original arterial phase images with the later delayed phase image sets also is needed. Although subtraction data sets provide visually appealing maximum intensity projection images for review with referring physicians, diagnosis of venous abnormalities is predominantly reliant on multiplanar reconstructions of the source data (arteriovenous images). In cases in which additional assessment of venous structures is needed after gadolinium-enhanced MR venography, a modified TOF MR sequence with appropriately structured saturation pulses can be used to suppress signal intensity after administration of the contrast material (24). With such a sequence, TOF MR venograms can be obtained to clarify or confirm 3D MR venographic findings, even after contrast material has been administered.
The major limitation of this study was the lack of correlation for all MR venographic results. Only three patients successfully underwent conventional venography for corroboration of MR venographic findings. Even in these three patients, MR venographic findings for veins not directly examined at conventional venography could not be confirmed. Similarly, in patients who underwent catheter or arteriovenous graft placement, the condition of veins that were not evaluated during these interventions could not be confirmed. Nevertheless, we have shown the feasibility of gadolinium-enhanced MR venography to provide appropriate guidance for medical management. It should also be noted that our study contained a relatively small number of patients. Therefore, additional studies will be necessary to confirm our results.
In conclusion, gadolinium-enhanced MR venography was able to help guide successful venous access or arteriovenous graft placement in all nine patients in whom such interventions were attempted and was predictive of one unsuccessful catheter placement. Although our study was predominantly focused on the evaluation of central veins in the thorax, the results of our preliminary experience (eg, in patient 4) suggest that the technique is similarly applicable to veins of the abdomen and pelvis. Three-dimensional gadolinium-enhanced MR venography has the potential to provide a rapid and comprehensive evaluation in patients in whom central venous access is necessary.
|
Footnotes |
|---|
Author contributions: Guarantor of integrity of entire study, T.S.S.; study concepts and design, T.S.S., V.S.L., N.M.R.; definition of intellectual content, T.S.S., V.S.L., N.M.R.; literature research, T.S.S.; clinical studies, V.S.L., N.M.R., G.A.K., J.C.W.; data acquisition, T.S.S.; data analysis, T.S.S., V.S.L.; manuscript preparation, T.S.S.; manuscript editing, V.S.L.; manuscript review, N.M.R., G.A.K., J.C.W.
|
References |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. Y. Kim, R. A. Mirza, J. A. Bryant, E. D. Whiting, D. M. Delong, C. E. Spritzer, and E. M. Merkle Central Veins of the Chest: Evaluation with Time-resolved MR Angiography Radiology, May 1, 2008; 247(2): 558 - 566. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Pedrosa, M. Morrin, L. Oleaga, J. Baptista, and N. M. Rofsky Is True FISP Imaging Reliable in the Evaluation of Venous Thrombosis? Am. J. Roentgenol., December 1, 2005; 185(6): 1632 - 1640. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H.-J. Baarslag, E. J.R. van Beek, M. M.W. Koopman, and J. A. Reekers Prospective Study of Color Duplex Ultrasonography Compared with Contrast Venography in Patients Suspected of Having Deep Venous Thrombosis of the Upper Extremities Ann Intern Med, June 18, 2002; 136(12): 865 - 872. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. J. Kroencke, M. Taupitz, R. Arnold, L. Fritsche, and B. Hamm Three-Dimensional Gadolinium-Enhanced Magnetic Resonance Venography in Suspected Thrombo-occlusive Disease of the Central Chest Veins Chest, November 1, 2001; 120(5): 1570 - 1576. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
