| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Clinical Observations |
1 Department of Radiology, Mater Misericordiae Hospital, Dublin, Ireland.
2 Department of Radiology, Cappagh National Orthopaedic Hospital, Cappagh
Hospital, Finglas, Dublin, Ireland 11.
3 Department of Oncology, Mater Misericordiae Hospital, Dublin, Ireland.
Received May 14, 2004;
accepted after revision October 15, 2004.
Address correspondence to S. J. Eustace.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CONCLUSION. Whole-body MRI represents an alternative to CT in the staging of lymphoma, with an ability to stage disease, identify lymph nodes greater than 1.2 cm, and the additional ability to evaluate for the presence or absence of disease spread to bone marrow. CT allows detection of more nodes (< 1.2 cm) than MRI but this does not alter tumor stage.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In contrast to CT, lymphadenopathy at MRI can be characterized on the basis of both size and signal characteristics. Similar to CT, size is used as a predictor of disease on MRI. However, in addition, changes in observed signal may improve reviewer confidence of tissue or nodal involvement. In this study, we compared nonionizing whole-body MRI to CT for the staging of lymphoma.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
CT Protocol
In each patient, CT was performed using a 4-MDCT scanner (Volume-Zoom,
Siemens Medical Solutions) with 4 x 2.5 mm acquisition from the skull
base to pubis. IV contrast material was used in seven patients who were
undergoing initial staging. Oral contrast material (barium sulfate suspension,
Readi-Cat; E-Z-EM) was used in all cases. In our department, we do not
routinely use IV contrast agents for follow-up lymphoma cases. After
acquisition, the images were transferred to a workstation (Magic View, Siemens
Medical Solutions). The images were reconstructed in 7-mm slices. If deemed
appropriate for interpretation, thinner slices were reconstructed in multiple
planes for definitive assessment. Seven patients did not have head and neck CT
examinations performed per the instructions of the referring oncologist. These
patients were still included in the study but had fewer individual sites
available for point-to-point comparison.
MRI
Each patient included for study underwent imaging on a 1.5-T Intera MR
(Philips Medical Systems) scanner using a moving tabletop and tabletop
extender, generating a longitudinal field of view of 200 cm and transverse
field of view of 53 cm.
In each case, images were acquired with a quadrature body coil with a moving tabletop technique in coronal and axial planes. Coronal scans were acquired in seven contiguous stations with 32 consecutive 8-mm slices at each station. The breath-hold technique was used in the thorax, with eight slices acquired for each breath-hold, allowing anteroposterior coverage in the coronal plane in four breath-holds. Images acquired in matching positions were automatically aligned to generate a seamless whole-body coronal image using prototypical software (View Forum software, Philips Medical Systems) and presented for interactive workstation review. Tissue excitation used turboSTIR with a TR of 4,257 msec, a TE of 58 msec, an inversion time of 150 msec, and an echo-train length of 40 with linear mapping of k-space. Axial images were acquired using a similar moving tabletop technique and turboSTIR tissue excitation.
After turboSTIR image acquisition in the coronal and axial planes, sagittal images of the spine were acquired using turbo spin-echo T1-weighted tissue excitation (TR/TE/turbo factor, 400/13/4). In three patients, axial gradient-refocused echo T1-weighted images were also acquired. A reconstructed voxel size of 1.03 x 1.02 x 8 mm could be achieved using the above methods.
Image Interpretation
To facilitate analysis of CT and MRI images, the body was divided into
lymph node or disease "stations," and at each station disease was
recorded as present or absent. If more than one lymph node was present at an
individual station, the size of the largest node was recorded. The size
recorded at CT was used as the gold standard for this study. Measurements were
recorded in short-axis dimension in one plane only in each case.
For the head and neck nodes, a modification of the system outlined by Som et al. [3] was used. By reference to readily identifiable anatomic markers on cross-sectional imaging, such as the hyoid bone, submandibular gland, mylohyoid, sternocleidomastoid, and anterior digastric muscles, lymph nodes could be placed into identifiable and defined stations that enabled accurate comparison between imaging techniques. A modification of the Lien and Lund [4] classification for the description of paratracheal and mediastinal nodes was used in the thorax. By this method, lymph nodes were recorded as present or absent in the anterior mediastinum, right paratracheal, right tracheobronchial, left paratracheal, left tracheobronchial, aortopulmonary window, subcarinal, and epicardial regions. If parenchymal nodules were identified, their lobar location was recorded.
In the abdomen and pelvis, record was made of paraaortic nodes, and if these were identified, their location in reference to an adjacent landmark such as a renal vessel was used to more accurately define its location. Similarly, record was made of mesenteric lymphadenopathy, iliac and pelvic sidewall adenopathy, inguinal lymphadenopathy, and visceral changes such as splenomegaly. Finally, evidence of bone marrow abnormality was sought by examining the CT images for sclerotic or destructive changes and the bone marrow for signal abnormality on MRI.
The CT images, having initially been interpreted by staff radiologists not involved in the study or in the interpretation of the whole-body MRI examinations, were then retrospectively reviewed by two study authors and the individual nodal stations were assessed. Multiplanar reconstructions were available but not used for this part of the study. The MR images were reported at the time of imaging to the referring oncologist. After a washout period of 3 months, the MR images were then reviewed for study purposes by the same two authors.
In all cases, the images were examined by the two radiologists in consensus, blinded to the result of the corresponding CT or MR image. This systematic evaluation for the presence of metastases led to the evaluation of at least 24 individual sites per patient and a further 14 sites if the head and neck were included. (For ease of interpretation, the head and neck lymph-node stations were not subdivided into A and B categories as proposed by Som et al. [3]). In total, allowing for the seven patients whose head and neck regions were not examined on CT, a total of 776 individual lymph node "stations" were assessed and available for comparison.
Statistics
The lymph node stations were divided by size criteria into three
categories: 1-6 mm, 7-12 mm, and > 12 mm. The sensitivity, specificity, and
positive and negative predictive values for MRI were calculated using CT as a
gold standard. In addition, by reference to clinical notes and bone marrow
histology results, the disease stage in individual patients was calculated
using both CT images and separately using MR images, and a direct comparison
in each patient was made.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Per Patient Analysis
An analysis of the accuracy of whole-body MRI in the assessment of
individual disease stage was also performed. Bone marrow biopsy results were
available in 18 patients. MRI correctly predicted marrow invasion in two
patients, thus correctly upstaging both patients to stage IV disease, and also
correctly predicted normal or noninvaded marrow in 16 patients. In all other
respects, the whole-body MRI and CT were concordant for disease staging. The
final stage by MRI: no evidence of disease, n = 14; stage
IE, n = 1; stage II, n =2; stage IIIX,
n = 4; and stage IV, n = 2. CT was unable to detect marrow
invasion in both cases, whereas MRI had an accuracy of 100%. The suffixes E
and X refer to extranodal and bulky, respectively. Bone marrow involvement is
predicted on whole-body MRI when extensive multifocal marrow signal
abnormality is present or when there is signal abnormality from places like
the epiphyses, which should contain no erythroid cell lines in adulthood
(Figs. 1A, and
1B).
|
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Coronal STIR tissue excitation involves a 180° prepulse with excitation of the returning longitudinal vector at the null point of fat [7]. Because it does not rely on frequency-selective suppression, it offers robust fat suppression over the whole body without the need for accurate shimming, and thus is useful for whole-body MRI. Lymphomatous tissue, like all neoplastic tissues, has increased water content that returns high signal on STIR imaging, enabling detection of disease [8-10] (Figs. 2A, 2B, 2C, and 2D). The acquisition of images in two planes also helped improve reviewer confidence (Figs. 3A, 3B, and 3C). In this study, we proved a high sensitivity for all disease over 12 mm, a size criterion we used to determine significance at CT. In fact, the only lesions over 12 mm missed by whole-body MRI were calcified lesions that were determined to be inactive at CT. One was a pulmonary nodule due to old mycobacterial disease and the other three were calcified mediastinal lymph nodes representing old, treated Hodgkin's disease. That whole-body STIR imaging did not detect these lesions is not surprising because a calcified lesion does not routinely return signal at MRI and would be largely undetectable against a fat-suppressed background. Axial T1-weighted imaging was initiated to offset this disadvantage, as the signal void returned from a calcified lesion might be easier detected against a bright background of fat.
|
|
|
|
|
|
|
Whole-body MRI as we use it is poor at detecting lung nodules and missed eight of nine nodules detected at CT, although these were all smaller than 12 mm. However, recent research [11-13] has shown that MRI can be very accurate at detecting small lung nodules. It is hoped that improvements in respiratory and cardiac gating may be used to allow detection of smaller nodules.
Another disadvantage of whole-body MRI is in the detection of root of mesentery nodes. We currently use the technique without oral contrast and without bowel paralytic agents. The combination of peristalsis and high signal from fluid in adjacent bowel loops makes detection of lymph nodes difficult, particularly on STIR imaging. Again, the addition of the axial T1-weighted image has proven useful here, as small root of mesentery nodes are more readily visible against the background of the bright mesenteric fat. In addition, the short acquisition time (approximately 20 sec for the abdomen) and the low signal return from intraluminal fluid make detection easier due to less peristalsis and competing signal from moving bowel. Similar to techniques that are used for MR enteroclysis and colonography, the addition of negative contrast agents may have a beneficial effect in this area on whole-body MRI.
The final disadvantage of whole-body MRI is cost and availability. The methods we use are not available on all systems, although most of the major vendors now offer some type of whole-body imaging capabilities. As MRI becomes more widely available, the cost of whole-body studies may lessen significantly. Whole-body MRI, although a relatively quick technique, is still relatively slow compared with MDCT. Gating and bowel preparation, which will improve diagnostic performance, will undoubtedly add time.
Whole-body MRI is extremely useful for detecting marrow signal abnormality and the addition of sagittal T1-weighted imaging further improves detection in the spine and sternum (Figs. 1A, and 1B). In all 18 cases where marrow results were obtained, whole-body MRI concurred with the available marrow biopsy results, which altered staging over CT staging alone in two cases. MRI has been shown as useful for the detection of marrow involvement by lymphoma [14-16], with high sensitivities and specificities. It would be logical to expect that whole-body MRI, which examines the patient's entire marrow, would have even higher accuracy than protocols that examine selected areas such as the pelvis and spine. Recent research suggests that MRI of marrow in hematopoietic malignancies may be improved by the administration of gadolinium [17] or iron-oxide particles [18], potentially further enhancing the accuracy of whole-body MRI for marrow invasion. Such strategies will, however, increase time and cost and remain to be validated. In particular, imaging of post-treatment marrow has been a subject of difficulty in the past; it is hoped that iron-oxide agents may help to differentiate between post-treatment change and recurrent disease.
The emergence of PET, particularly PET/CT, has had a profound influence on oncology over the past few years. There is an accumulating body of evidence that this will become the new gold standard for imaging of lymphoma. Comparative studies suggest its superiority over CT [19] alone and further advantages in differentiating sterile tumor masses from viable tissue [20]. In addition, PET/CT seems capable of predicting response to treatment [21]. PET does have limitations, particularly in relation to physiologic uptake in the brain, heart, salivary glands, and kidneys [22]. Although coregistered images help to accurately determine the site of uptake, SUV values between lymphomatous and physiologic uptakes can overlap. This is particularly true of low-grade lymphomas in which PET has a less clearly defined role [23]. A major disadvantage of PET/CT is the dose of radiation incurred by the patient, which can lead to substantial cumulative doses to radiosensitive organs. Whole-body mean effective dose from FDG on an average-weight basis has been calculated at approximately 10.73 mSv for a 370 MBq dose [24]. The addition of a whole-body CT scan significantly adds to this, even if low-dose techniques are used. A recent abstract documented an effective dose of 19.262 mSv for a whole-body PET/CT scan using 370 MBq [25]. Thus, a nonionizing alternative imaging technique that has similar whole-body capabilities would be attractive. Recent evidence in the radiology literature has emphasized the potential risks of ionizing radiation in a screening population [26], and the doses from PET/CT are higher than those quoted. In addition, many patients in whom first-line chemotherapy has failed receive therapeutic radiation that further increases patient exposure. Radiosensitive organs such as the thyroid gland and breasts are especially vulnerable to the effects of radiation, and several studies have shown increased risks of thyroid dysfunction [27] and breast carcinoma [28] in this patient population. It is reasonable to assume that the effects of repeated imaging tests would be synergistic with radiation therapy, thus adding incentive for developing a nonionizing method of disease surveillance.
In conclusion, we report on our initial experience with whole-body MRI for the staging of lymphoma. In this study, whole-body MRI compared favorably with CT for nodes larger than 12 mm in short-axis diameter. Further refinements in technique, including the use of contrast agents, may further improve the technique's accuracy. PET/CT has recently emerged as a gold-standard test, and comparative tests between these two emerging techniques would be useful. For young patients in particular, in whom extensive longitudinal follow-up is anticipated, whole-body MRI may offer an alternative nonionizing method of staging and disease surveillance.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  R. C. Domingues, M. P. Carneiro, F. C. R. Lopes, R. C. Domingues, L. M. B. da Fonseca, and E. L. Gasparetto Whole-Body MRI and FDG PET Fused Images for Evaluation of Patients with Cancer Am. J. Roentgenol., April 1, 2009; 192(4): 1012 - 1020. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Brauck, M. O. Zenge, F. M. Vogt, H. H. Quick, F. Stock, T. Trarbach, M. E. Ladd, and J. Barkhausen Feasibility of Whole-Body MR with T2- and T1-weighted Real-time Steady-State Free Precession Sequences during Continuous Table Movement to Depict Metastases Radiology, March 1, 2008; 246(3): 910 - 916. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. C. Kwee, R. M. Kwee, and R. A. J. Nievelstein Imaging in staging of malignant lymphoma: a systematic review Blood, January 15, 2008; 111(2): 504 - 516. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Muntanola, F. Bosch, P. Arguis, E. Arellano-Rodrigo, C. Ayuso, E. Gine, M. Crespo, P. Abrisqueta, C. Moreno, F. Cobo, et al. Abdominal Computed Tomography Predicts Progression in Patients With Rai Stage 0 Chronic Lymphocytic Leukemia J. Clin. Oncol., April 20, 2007; 25(12): 1576 - 1580. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
