| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Editorials |
1 From the Weill Medical College of Cornell University, 416 E 55th St, New York, NY 10022. Received November 12, 2002; accepted November 14. Address correspondence to M.R.P. (e-mail: map2008@med.cornell.edu).
Index terms: Editorials • Magnetic resonance (MR), utilization • Veins, MR, 93.12942, 94.12942, 98.12942
I am not in the giving vein . . .
William Shakespeare (1)
Radiation exposure from x-ray studies creates risks that are well within what we tolerate as part of day-to-day activities of routine life. For example, the risk of death from radiation exposure from abdominal computed tomography (CT) is purported to be lower than the risk of cancer death from smoking cigarettes for 1 year (2). However, despite controversy regarding the health effects of low-dose ionizing radiation and because of the belief that there are finite risks, extra effort is expended, and mandated by regulatory authorities, to minimize the magnitude of radiation exposure that occurs during a single diagnostic imaging examination. For example, fluoroscopy is carefully monitored so that high-dose exposures occur only at opportune moments at which anatomy or disease is shown to best advantage. The ALARA (as low as reasonably achievable) principle for radiation exposures requires vigilant attention on the part of the radiologist to ensure the proper balance between more radiation to improve quality or less radiation when quality is already sufficient to resolve the clinical issues.
But with magnetic resonance (MR) imaging, there is no ionizing radiation exposure. This opens up the possibility of continuously imaging over longer periods of time than are reasonable with modalities in which ionizing radiation is used. One can image over an entire physiologic cycle and later select the optimum data for the final highest-quality image to make the diagnosis or to illustrate the diagnosis to referring physicians. Continuous imaging with subsequent postprocessing eliminates the need for a radiologist to be present to monitor image data acquisition. There may even be less need to anticipate or determine the purpose of the examination, because the acquired data may be manipulated to address a broader array of indications. This offers the potential to migrate to more efficient use of radiologist and technologist time.
For vascular studies, an important cycle is the transit of a contrast material bolus through the anatomy of interest. Continuous imaging in which multiple transit phases of a contrast material bolus are acquired eliminates the need for contrast material injection timing. Because data are always being acquired, improper bolus timing is no longer possible and is eliminated as a cause of failed examinations.
Of course, radiologist supervision of imaging is not eliminated but is rather shifted from careful monitoring of data acquisition to careful supervision of the subsequent image processing. Real-time monitoring is replaced by postprocessing. Postprocessing may be as simple as selecting the best image, or it may involve averaging, subtracting, filtering, or other techniques for improving signal-to-noise ratio or contrast or eliminating artifact. When data are three-dimensional (3D), postprocessing can also involve an array of projection and volume rendering techniques to transform 3D data into two-dimensional (2D) images for display. Clever postprocessing routines may create images that never actually exist but nevertheless show the desired structure or physiologic alteration to best advantage for diagnosing disease. In particular, spatial subtraction (removing overlying structures) reveals hidden structures, and temporal subtraction (removing superimposed temporal changes) reveals hidden pathophysiology.
One such application for venous imaging is presented in the current issue of Radiology by Fraser et al (3). Eight series of 3D MR angiographic data are acquired during bolus injection of a gadolinium-based contrast material. A precontrast data set and an arterial phase data set are subtracted from two late (equilibrium phase) data sets to create a final 3D image that never actually existed: a pure venogram. Normally, at conventional venography, a pure venogram can only be obtained by cannulating a dorsal foot vein and performing sequential imaging during injection of iodinated contrast material. MR venography can also be performed by means of sequential imaging during injection of dilute gadolinium-based contrast material via a dorsal foot vein (4). However, the venous enhanced subtracted peak arterial (VESPA) approach eliminates the need to cannulate a foot vein, eliminates the need to use ionizing radiation, and eliminates the problem of the nephrotoxicity of iodinated contrast material. Intravenous injection of a gadolinium chelate is still necessary, but an antecubital approach is acceptable.
Of course, MR venography has been possible for years without gadolinium or postprocessing. Simple protocols that incorporate 2D time-of-flight acquisitions are highly accurate for the diagnosis of deep venous thrombosis (DVT) involving the femoral, popliteal, or iliac veins. MR venography is the best technique for evaluation of gonadal vein thrombosis (5). But despite the advantages of high accuracy and noninvasiveness, and the limited utility of ultrasonography (US) for clinical scenarios other than symptomatic femoropopliteal DVT, 2D time-of-flight MR venography is not commonly used. Will the VESPA technique, with its shorter examination time and 3D venous images, allow MR venography to play a greater role in routine diagnostic evaluation of iliofemoral DVT?
There are two categories of issues that need to be considered when envisioning the application and development of this technique. First, there are clinical and pathophysiologic factors that will influence the practicality of the examination, its suitability for patient diagnosis (given various clinical scenarios), and whether physicians will use it in practice. Second, there are technical issues that will affect the performance, practicality, and future improvements of this method.
From the clinical perspective, we need first to consider why unenhanced 2D time-of-flight MR venography, despite its high accuracy and complete noninvasiveness, has not been widely adopted for the evaluation of suspected DVT, while compression US remains the standard despite its well-documented limitations, which include inability to enable reliable evaluation of pelvic veins and low sensitivity in patients without clinically localizing findings in the legs (6,7). The answer may be that MR venography is simply less practical for the busy clinician to use. It is less easily performed emergently than US, issues with monitoring and claustrophobia persist (particularly in dyspneic patients who may also have pulmonary embolism [PE]), and MR imaging is perceived as substantially more expensive than US. Accordingly, US may be viewed as "good enough" by the referring physician despite its limitations, much as ventilation-perfusion scanning was used routinely until a more comprehensive and accessible examination (CT) became available. It will be hard for VESPA or any other technique to improve on the already excellent accuracy of unenhanced 2D time-of-flight acquisitions, and, while VESPA images may be more appealing and persuasive than 2D time-of-flight sections, it is not clear that their appeal will overcome the disincentives for referring physicians to use MR imaging in preference to US for the diagnosis or exclusion of DVT.
Another major clinical issue is that venous thromboembolism is a systemic (or at least multisite) disorder, of which PE and DVT are individual manifestations. Therefore, a single examination that could enable imaging of both the pulmonary arteries and the pelvic and lower extremity veins would change the clinical paradigm and (probably) referring physician behavior. The pulmonary arteries can be visualized at MR imaging, but, although the results of several studies (8–10) have been excellent, this technique has not been widely adopted. CT has become the main noninvasive modality for directly visualizing the pulmonary arteries and PE. Radiation and the need to use iodinated contrast material are drawbacks of this modality, but they have not dissuaded referring physicians from adopting it widely, even in the face of conflicting data regarding its accuracy.
Recently, some investigators have combined CT pulmonary angiography with venous phase imaging of the lower extremity and pelvic veins (so-called indirect CT venography) (11–13). The indirect CT venography technique involves an increase in the ionizing radiation dose but takes advantage of the iodinated contrast material used for CT pulmonary angiography so there is no incremental risk of nephrotoxicity. A recent review (11) and two large multicenter clinical studies (12,13) of CT pulmonary angiography/indirect CT venography revealed excellent correlation with US and discovery of clinically important unsuspected DVT that warranted alteration in therapy plans. Thus, the competition for VESPA MR venography may be CT rather than US.
Can VESPA be used in an analogous protocol that would combine MR venography with MR pulmonary angiography? Given that both arterial phase imaging and venous phase imaging of the lower extremities are required, it seems unlikely, unless (a) the radiologist is prepared to administer a second injection of gadolinium chelate for pulmonary artery imaging or (b) simultaneous multisite surface-coil imaging techniques with high spatial and temporal resolution can be developed that will encompass the thorax, pelvis, and lower extremities—a nontrivial challenge for radio-frequency and gradient technology. Should the technical obstacles be overcome, MR imaging investigators would still need to show their referring colleagues data demonstrating that MR pulmonary angiography is accurate, technically robust, and tolerated by sick patients. Thus, one wonders if a new MR venography technique like VESPA, burdened with additional clinical requirements beyond its initial technical scope, will succeed in the clinical environment.
One interim answer may lie in "niche" applications. The VESPA MR technique does offer an improved solution for an area of medicine that is currently not well addressed by any imaging technique other than the relatively slow 2D time-of-flight MR technique. US is limited for evaluation of iliac veins and the distal portion of the inferior vena cava, as well as abdominal and upper extremity veins and the superior vena cava. With the high spatial resolution conferred by VESPA, pure venographic images without confusion of overlapping arteries can be obtained in a simple, timing-independent examination. If VESPA or similar techniques gain a foothold in specialized clinical scenarios, their data quality could become increasingly persuasive for their use for other clinical indications.
There are also important technical issues to consider. Most important is the opportunity to automate data acquisition. Elimination of the need for a radiologist to monitor pulmonary arterial CT scanning has made CT more accessible and popular with referring physicians. To the extent that the use of VESPA simplifies and automates MR data acquisition by eliminating complexity and the need for radiologist monitoring, it may attain the same benefit for MR. However, it will be necessary to automate postprocessing as well. This improvement in postprocessing will no doubt require faster workstations with larger memory banks and new, more intuitive software. Technologists may need to have more emphasis in their training on postprocessing as well as workstation trouble shooting.
Another technical challenge is the need for patients to hold perfectly still for the entire period of data acquisition to avoid misregistration between phases. With the VESPA technique this problem is solved by using a vacuum cushion to immobilize the legs. Ultimately, the period of time that a patient can lie motionless will limit the number of useful phases that can be acquired for postprocessing. Some motions such as cardiac pulsation, respiration, and peristalsis cannot be easily prevented and will require more sophisticated approaches, such as the use of navigators, which are still being developed.
One exciting potential for improvement relates to MR contrast agents. The current United States Food and Drug Administration–approved agents are avidly extracted by most tissues, resulting in a short vascular half-life. Blood pool contrast agents that are now being developed avoid extraction by tissues and thereby attain higher vascular signal-to-noise ratio during equilibrium phase imaging. The longer vascular half-life of blood pool contrast agents will allow more opportunity for increased signal averaging. Use of blood pool contrast agents may allow the imaging of additional sites, such as the pulmonary arteries, that do not require arterial phase imaging. This would then provide the opportunity to change physician behavior by making the MR imaging examination for thromboembolic disease more comprehensive. Blood pool agents also tend to have several-fold higher T1 relaxivity. This offers the opportunity to inject the equivalent of 10 mL/sec of a standard gadolinium chelate contrast agent at a more reasonable rate that is several-fold lower. The higher equivalent injection rate produces a higher concentration bolus, thereby improving arterial phase imaging as well. Improving both the arterial phase and equilibrium phase images will presumably make the final pure venogram superior.
Eliminating nephrotoxicity and radiation exposure is a major improvement, but it does not guarantee adoption of a new technique. Simplification, convenience for referring physicians, comprehensiveness, and efficient use of radiologist time are also critical to the success of new techniques.
FOOTNOTES
See also the article by Fraser et al in this issue.
REFERENCES
Related Article
This article has been cited by other articles:
|
|
 |
|
  H. H. Hu, N. G. Campeau, J. Huston III, D. G. Kruger, C. R. Haider, and S. J. Riederer High-Spatial-Resolution Contrast-enhanced MR Angiography of the Intracranial Venous System with Fourfold Accelerated Two-dimensional Sensitivity Encoding Radiology, June 1, 2007; 243(3): 853 - 861. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Obernosterer, M. Aschauer, H. Portugaller, H. Koppel, and R. W. Lipp Three-Dimensional Gadolinium-Enhanced Magnetic Resonance Angiography Used as a "One-stop Shop" Imaging Procedure for Venous Thromboembolism: A Pilot Study Angiology, July 1, 2005; 56(4): 423 - 430. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
