| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Commentary |
1 Department of Radiology, University of California, San Francisco, 505 Parnassus Ave., Suite M392, San Francisco, CA 94143-0628.
Received November 19, 2006; revised November 30, 2006;
Address correspondence to L. S. Steinbach
(lynne.steinbach{at}radiology.ucsf.edu).
Keywords: bone • MRI • oncology
As a historical note, we originally presented this paper at the RSNA annual meeting in 1984, where it was given in a session on nuclear medicine! We had difficulty getting the paper published in AJR because the reviewers insisted we should have been able to distinguish different types of neoplasms by adjusting the pulse sequences. Of course, we now know that is nonsense. The point we were trying to make was that we wanted to identify abnormal marrow areas that could then undergo successful biopsy. To that end, we were successful.—Richard Daffner
During the early 1980s, radiology departments were just beginning to experiment with MRI for the evaluation of disorders of the brain. In 1985, MRI was just getting off the ground in the area of musculoskeletal imaging. We realized that we could look at meniscal tears, rotator cuff tears, and other joint derangements. We could also visualize and characterize soft-tissue masses. At that time, we were not sure if MRI would be useful for evaluating bone marrow, and there were statements made to the contrary. It did not take long to discover that MRI had the unique capability to evaluate medullary bone.
One of the first articles to discuss the use of MRI for detection of metastatic disease and multiple myeloma in the bone marrow was published in 1986. Richard (Dick) Daffner and colleagues [1] from Allegheny Hospital in Pittsburgh, Pennsylvania, wrote this landmark article that is included as one of the most influential articles published in the American Journal of Roentgenology during its first 100 years [1]. They evaluated the use of MRI for detection of disease in the bone marrow of 80 patients with known metastatic disease of the skeleton or with multiple myeloma. In those days, all radiologists were imaging at lower field strengths, and this study used a 0.5-T MR unit (Magnetom, Siemens Medical Solutions) operating at 0.35 T. All lesions were biopsy proven. Fifty patients had a malignant primary source of metastases and 30 patients had multiple myeloma. The MR examinations were correlated with radiographs and bone scans. The authors noted that the bone marrow showed low signal intensity on T1-weighted images in 40 patients with metastases and in all patients with myeloma. Only 10 patients did not show an abnormality on T1-weighted images. In contrast to the sensitivity of MRI, most of the bone scans of the patients with multiple myeloma showed normal results.
Around the time of submission of the article by Daffner et al. [1], a few articles came out in the AJR concerning MRI of bone marrow, including one by Wismer et al. [2] on chemical shift imaging of bone marrow and another by Ramsey and Zacharias [3] regarding MRI of the spine after radiation therapy. Other journals were not yet publishing articles on MRI of the bone marrow. Pulse sequences were being explored, and STIR imaging and other variants of T2-weighted imaging were being recommended for further characterization of the lesions. Shortly thereafter, there was an explosion in the use of MRI for evaluating malignancy in the bone marrow.
This was especially true in the area of leukemia. Subsequently, researchers showed good results for the use of MRI for leukemia and aplastic anemia before and after treatment [4, 5]. The bone marrow could undergo a flip-flop pattern of signal intensity in pediatric patients with cancer [6]. Analysis of MRI for diagnosis and treatment of multiple myeloma was further investigated [7, 8]. MRI was found to be sensitive for identifying abnormalities of the marrow in a small number of patients with other monoclonal gammopathies [9]. Evaluation of marrow response to various forms of cancer therapy on MRI was also addressed [10]. Pitfalls that might mimic malignancy, such as hematopoietic hyperplasia, which caused a few unnecessary biopsies in the early days, were elucidated and ways to distinguish malignancy from abnormal marrow were suggested [11]. In the present day, we realize that this distinction is not always a perfect science because hyperplastic marrow can mimic tumor and enhance with gadolinium.
New techniques were tried for distinguishing between malignant and benign bone marrow with varying success. These included T1–T2 ratios [12], quantitative chemical shift imaging [13–16], and the use of fat suppression for increasing the dynamic range allowing for increased conspicuity of lesions [17]. Fast spin-echo imaging was faster than STIR imaging and allowed for improved tissue specificity [18]. Using moving tables and special coils to survey the whole body, whole-body MRI was investigated for screening to search for primary tumors and metastases with much success [19]. This practice is limited by equipment and expertise but is currently used at some centers. The administration of dynamic IV gadolinium for osseous malignancy was also investigated for increasing specificity but has not shown utility in clinical practice [20]. Diffusion imaging has been investigated as a way to gain increasing accuracy for the diagnosis of malignancy [21, 22], but it has not been widely adopted by radiologists because of surrounding controversy regarding specificity for malignancy.
As enthusiastic as we are to find a test that will diagnose a lesion with certainty, similar to creating a microscopic look at the bone marrow, we realize that MRI has its limitations. Certain characteristics could distinguish benign from malignant compression fractures [23], although not all of the time. The application of MRI for evaluation of multiple myeloma has met with varying success [24]. MRI is currently used for staging and treatment monitoring of multiple myeloma in some centers because it is more sensitive than radiographic bone surveys or bone scintigraphy for the detection of marrow lesions. Other centers do not use MRI routinely for myeloma. The use of MRI to predict response duration and survival from bone marrow transplantation for multiple myeloma has been investigated [25].
In a study of 29 patients with chronic lymphocytic leukemia, quantitative MRI failed to detect leukemic marrow infiltration in 41% [26]. Currently, there is no role for MR evaluation of bone marrow in leukemic patients; however, MRI can answer specific questions regarding bone marrow biopsy sites when there is a negative bone marrow aspirate, and it can also detect complications of treatment such as osteonecrosis. We have found MRI very sensitive but less specific for metastatic lesions. Posttreatment changes and therapies for bone marrow regeneration could affect the marrow, may simulate malignancy, and are nonspecific [27].
Realizing the utility of MRI and accepting some of the lack of specificity, the current quest is to find the test that will be complementary to MRI and, at times, even more sensitive and specific for metastatic disease, myeloma, and other malignancies of the bone marrow. Bone scintigraphy is still used as the mainstay for detecting metastatic disease in the marrow. PET is useful for detection of multiple myeloma [28]. Specific marrow contrast agents in the category of ultrasmall superparamagnetic iron oxide particles show promise for identifying marrow-occupying lesions [29]. For all new imaging techniques, the literature will probably recycle through a similar pathway of exploration and reflection as has been the case with MRI for malignant bone marrow disorders.
References
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  M. D. Cohen MRI and the Detection of Bone Marrow Tumor Am. J. Roentgenol., January 1, 2008; 190(1): W75 - W75. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
