| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Editorials |
1 From the Department of Radiology, Penn State Milton S. Hershey Medical Center, PO Box 850, 500 University Dr, Rm M108, Mail Code H066, Hershey, PA 17033. Received April 26, 2002; accepted May 3. Address correspondence to the author (e-mail: tmosher@psu.edu.
Index terms: Bone marrow • Editorials • Gadolinium • Magnetic resonance (MR), contrast enhancement, 30.12141, 40.12141 • Special Communications
The more times you run over a dead cat, the flatter it gets.
McClure Wilson (1)
A continuing challenge to the radiologist is managing the uncertainty of clinical diagnosis. A natural response when faced with an equivocal finding is to obtain more information. Ideally, this information provides novel insight into the question at hand; however, information obtained is often redundant and simply confirms knowledge already possessed.
A frequent diagnostic dilemma in musculoskeletal magnetic resonance (MR) imaging is the determination of the cause of abnormal bone marrow signal intensity. This is especially common in studies of the foot and ankle where abnormal marrow signal intensity is prevalent. A standard MR imaging protocol for bone marrow evaluation is a combination of T1-weighted and T2-weighted fat-suppressed imaging with either short inversion time inversion-recovery (STIR) or T2-weighted fat-suppressed fast spin-echo (SE) techniques.
Unfortunately, although imaging with STIR and T2-weighted techniques is very sensitive, findings on images obtained with them are not specific. A variety of pathologic conditions, which include edema, hemorrhage, necrosis, microfracture, infection, neoplasm, and even biomechanical stress from exercise, can produce similar abnormal findings (2). Despite a lack of data to support the practice of performing additional contrast material–enhanced studies when there are equivocal findings of abnormal marrow signal intensity, many radiologists do so. Although this increases the length and expense of the MR examination, it is not clear if this practice improves diagnosis.
In this issue of Radiology, Schmid et al (3) report results of a prospective study comparing turbo STIR and T1-weighted contrast-enhanced fat-suppressed turbo SE MR imaging in diagnostic interpretation of bone marrow abnormalities in the foot and ankle. The authors systematically evaluate three parameters that potentially influence the diagnostic thinking of the radiologist: (a) contrast, (b) volume, and (c) pattern of abnormal bone marrow signal intensity on STIR and T1-weighted gadolinium-enhanced fat-suppressed MR images.
Although contrast was greater on T1-weighted fat-suppressed MR images, there was no difference in volume or pattern of abnormal bone marrow signal intensity. More importantly, there was very little difference in diagnostic interpretation of this finding. Although these researchers did not attempt to determine diagnostic accuracy, the high level of agreement in radiologic diagnosis implies equivalent accuracy of the two techniques. The authors conclude that addition of T1-weighted contrast-enhanced MR imaging does not alter diagnosis, and for most cases, they recommend performing only the STIR sequence.
Although the benefit of using gadolinium-based contrast material in MR imaging of the brain, spine, and cardiovascular system is well established, indications for use of this contrast material in musculoskeletal MR imaging remain controversial. Addition of gadolinium-based contrast material has the potential to increase lesion detection and provides a mechanism for improved lesion characterization. Increased sensitivity of gadolinium-enhanced MR imaging is clearly established in imaging of the neurologic system, where the blood-brain barrier provides a physiologic basis for establishing contrast between normal and abnormal tissue. However, no equivalent barrier exists for the musculoskeletal system. Here, contrast is a function of temporal differences in biodistribution of gadolinium-based contrast material between normal and abnormal tissue.
In theory, the results of Schmid and colleagues (3) are not unexpected. Because gadopentetate dimeglumine distributes into free extracellular water, increased signal intensity on contrast-enhanced equilibrium-phase MR images reflects increased extracellular water content of tissue. Since increased water content is highly correlated with longer T1 and T2 times, it will increase signal intensity on images obtained with the STIR technique. Both gadolinium enhancement and increased signal intensity on images obtained with the STIR sequence reflect similar differences in water content of tissue. Dynamic imaging immediately after administration of a bolus of a gadolinium-based contrast agent provides image contrast that is based on tissue capillary density and vascular permeability. This information can be used to identify pathologic processes with a high level of biologic activity, such as acute inflammation, infection, or malignancy. However, since hyperemia and increased vascular permeability are associated with higher interstitial water content, it is not safe to assume that the results of Schmid and colleagues would have been substantially different had they used dynamic rather than equilibrium-phase imaging.
In their study, Schmid and co-workers (3) did not detect any additional bone marrow abnormalities on T1-weighted contrast-enhanced fat-suppressed MR images. This result supports prior conclusions by researchers (4) who reported a similar sensitivity with use of STIR and T1-weighted gadolinium-enhanced fat-suppressed MR imaging in detection of bone marrow disorders. Is there an indication for use of gadolinium-based contrast material for the detection of abnormalities in bone marrow signal intensity? One indication may be for use in the evaluation of small joints, or another indication may be for use in cases requiring high spatial resolution. Although, when necessary, T2-weighted images with a higher signal-to-noise ratio (SNR) can be obtained by using a T2-weighted fat-suppressed fast SE sequence rather than a STIR sequence, both have a low SNR compared with that of sequences with a short echo time.
In certain cases, the T1-shortening effect of gadolinium, and the subsequent increase in image SNR, may have a diagnostic benefit, such as in the detection of small cortical erosions (5), early osteonecrosis (6), or periosteal edema (7). The lack of contrast enhancement can be beneficial for the detection of small regions of avascular tissue, such as cystic necrosis, a cyst, or an abscess. In this situation, dynamic imaging should be performed to minimize the effect of diffusion of contrast material into the fluid compartment. In the study of Schmid et al (3), of the nine cystlike lesions identified on images obtained with the STIR sequence, two (22%) lesions were not characterized as cysts on the contrast-enhanced images by one of the readers. For this particular pattern of abnormal bone marrow signal intensity, dynamic imaging could potentially improve diagnostic accuracy. Dynamic imaging may also be useful in detection of small hypervascular lesions, such as the nidus of osteoid osteomas that may be obscured by surrounding marrow edema on images obtained with the STIR or T2-weighted fat-suppressed sequences.
More often, in clinical practice, the difficulty is not in the detection of bone marrow abnormalities but in the confident diagnosis of the cause. Does gadolinium-based enhancement improve specificity in the diagnosis of bone marrow abnormalities? In evaluation of the foot and ankle, the conclusion of Schmid and colleagues (3) is that it does not. There is reason to believe that this conclusion is true for other regions of the body. Although underlying causes frequently differ, the finding of nonspecific bone marrow edema often is observed in the proximal femur. In this region as well, there is likely a strong correlation between findings on images obtained with gadolinium enhancement and on those obtained with the STIR sequence or on those obtained with the T2-weighted fat-suppressed sequence.
In their discussion, Schmid and colleagues (3) are careful not to extrapolate their findings to evaluation of neoplasms or infection. Evaluation of the pattern of contrast enhancement has been useful in helping to distinguish an acute infarct from osteomyelitis in patients with sickle cell disease (8). In osteomyelitis, the lack of enhancement may be useful for identification of regions of devitalized bone. However, in the evaluation of osteomyelitis of the foot in a diabetic patient, contrast-enhanced MR imaging does not help to differentiate osteomyelitis from bone marrow edema (9). Dynamic enhanced imaging of bone marrow tumors can reveal regional differences in biological activity of the tumor that can be useful in planning before biopsy, identification of tumor necrosis, and monitoring of the response to therapy (4,10).
However, contrast enhancement has not been found to result in marked improvement for differentiation between benign and malignant neoplasms or for determination of tumor margins. In general, for depiction in the study of patients with infection and neoplasm, the primary benefit of contrast enhancement is in the evaluation of adjacent soft tissues (10,11). Carefully controlled clinical trials would be useful for determination of the diagnostic effectiveness of contrast-enhanced MR imaging in patients with a musculoskeletal neoplasm or infection. These studies may help to identify specific indications and to determine the role of dynamic contrast-enhanced imaging.
A potential benefit of gadolinium enhancement is increased diagnostic confidence. This parameter was not evaluated in the study of Schmid and colleagues (3). For small lesions, such as subtle subcortical insufficiency fractures or osteonecrosis, the greater SNR and contrast resolution of T1-weighted gadolinium-enhanced fat-suppressed images may provide a clinically important improvement in diagnostic confidence that would have prognostic implications for the patient.
In a study regarding the evaluation of subchondral marrow edema of the femoral head to determine prognostic features for differentiation between irreversible and transient lesions, Vande Berg et al (12) found no substantial difference between findings with contrast-enhanced and with T2-weighted images. However, lesions were found to be more conspicuous on enhanced images, which helped because it augmented confidence in detection. In clinical practice, there will be situations in which additional confirmation provided by enhanced images is necessary, even though the presumed diagnosis is not changed. However, such cases are likely the exception rather than the rule.
A certain level of diagnostic uncertainty is inevitable, and for some cases a diagnosis is unnecessary. Occasionally, the objective of the MR imaging study is not diagnosis but rather prognosis. Few researchers have evaluated the clinical outcome of patients with abnormal bone marrow signal of the foot and ankle. In a recent study, Zanetti et al (13) evaluated 31 patients with edema-like marrow abnormalities of the foot and determined that 179 (54%) had persistent pain 1 year after the MR examination. The pattern of marrow abnormality influenced prognosis. Patients with well-defined necrosis-like lesions were most likely to have persistent pain, whereas patients with probable stress fractures were most likely to be pain free.
Stress-related marrow edema is likely to be transient and self-limiting with conservative treatment, whereas osteonecrosis is more likely to lead to long-term symptoms. In patients with rheumatoid arthritis, identification of an active inflammatory pannus with contrast-enhanced MR imaging had prognostic value in the prediction of progression of further bone erosion (14). Additional clinical trials in which researchers determine if specific patterns of abnormal marrow signal intensity or enhancement predict therapeutic outcome would be especially helpful, as they may impact patient treatment.
Schmid and colleagues (3) are to be commended for applying specific quantifiable criteria in their evaluation. Their results illustrate a valuable role for technology assessment in clinical practice. In rapidly developing fields, such as MR imaging, there is a tendency to accept new techniques prior to thorough evaluation of their technical or diagnostic performance. Acceptance of new methods often is based primarily on aesthetic or personal preference rather than on sound scientific evaluation. With MR imaging, the availability of different forms of contrast provided by different pulse sequences can give the illusion that the resultant images provide different diagnostic information. However, often the observed differences in contrast reflect the same underlying pathophysiologic finding.
The fine line between confirming a diagnosis and performing an unnecessary procedure can vary on the basis of the specific case and the experience of the radiologist. Therefore, the conclusions of Schmid and colleagues (3) should not be broadly extrapolated to define strict guidelines for patterns of practice. However, their results should cause us to question the need for use of gadolinium-based contrast material when faced with the diagnostic challenge of abnormal bone marrow signal. In the climate of cost containment, use of contrast enhancement with MR sequences must, at a minimum, have the potential to alter diagnosis. The decision to use contrast material should be directed toward answering a specific diagnostic question that cannot be answered through simpler methods.
It has been the experience of this author that, when faced with the finding of nonspecific marrow abnormalities of the foot and ankle, the most useful information is usually gathered from a review of the patient’s history and from findings at physical examination. When faced with diagnostic uncertainty in the evaluation of abnormal marrow findings, indiscriminant use of contrast-enhanced images simply serves to confirm the obvious, with little diagnostic effectiveness. As in words attributed to Dr McClure Wilson through which he describes the futility of repetitive diagnostic testing, "The more times you run over a dead cat, the flatter it gets" (1).
FOOTNOTES
See also the article by Schmid et al in this issue.
REFERENCES
Related Article
This article has been cited by other articles:
|
|
 |
|
  J. L. Bloem, T. J. Mosher, M. R. Schmid, M. Zanetti, H. P. Ledermann, W. B. Morrison, and M. E. Schweitzer Dynamic Gadolinium-enhanced MR Imaging in Bone Marrow Disorders [letter] * Dr Mosher responds: * Drs Schmid and Zanetti respond: * Dr Ledermann and colleagues respond: Radiology, April 1, 2003; 227(1): 303 - 305. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
