Skeletal Radio1 (1991) 20:79-84
Skeletal Radiology
Magnetic resonance imaging of disseminated bone marrow disease in patients treated for malignancy S.L. Hanna, M.D., 1 B.D. Fletcher, M.D., 1 D.L. Fairclough, Dr. p.j.,2 J.H. Jenkins III, M.D., 3 A.H. Le, B.S. 1 Departments of 1Diagnostic Imaging, 2Biostatistics and Information Systems, and 3Pathology and Laboratory Medicine, St. Jude Children's Research Hospital and The University of Tennessee, Memphis, Tennessee, USA
Abstract. Magnetic resonance imaging ( M R I ) is a sensitive method for the diagnosis of bone m a r r o w abnormalities, but its usefulness in detecting active disseminated cancer in this tissue in treated patients has not been determined. We therefore examined 14 children who had been treated for disseminated bone m a r r o w involvement by neuroblastoma ( n = 6 ) , l y m p h o m a ( n = 3 ) , Ewing's sarcoma (n = 3), osteosarcoma (n = 1), and leukemia (n = 1). M R I studies were performed at 21 m a r r o w sites to evaluate residual or recurrent t u m o r and were correlated with histologic material from the same site. T1- and T2weighted sequences were employed in 21 and 14 studies, respectively; short tau inversion recovery (STIR) in 18; and static gadolinium diethylene triamine pentaacetic acid (Gd-DPTA)-enhanced, T~-weighted sequences in 13. All M R I studies showed an altered bone m a r r o w signal. Technetium 99m methylene diphosphonate (99mTc-MDP) bone scintigraphy was also performed (19 studies). On histologic examination, 7 m a r r o w specimens contained tumor, and 14 did not. O f the 7 tumor-positive lesions, all Ta-weighted, 4 of 6 T/-weighted, and all 6 S T I R sequences showed a b n o r m a l signal; all 5 GdDTPA-enhanced, Tl-weighted sequences showed enhancement of the lesion. However, abnormal signals were also observed on all Tx-weighted, 6 of 8 Tzweighted, 11 of 12 STIR, and 5 of 8 G d - D T P A - e n hanced, T~-weighted images of the tumor-negative sites. In this clinical setting, M R I did not consistently differentiate changes associated with treatment from malignant disease. Key words: Bone marrow, magnetic resonance studies - Bone marrow, biopsy - Magnetic resonance c o m p a r a tive studies Magnetic resonance in infants and children - Neoplasms
Address reprint requests to. Barry D. Fletcher, M.D., Department
of Diagnostic Imaging, St. Jude Children's Research Hospital, P.O. Box 318, Memphis, TN 38101, USA
Magnetic resonance imaging ( M R I ) is a valuable method for demonstrating malignant bone m a r r o w infiltration because of the excellent contrast between abnormal tissue and normal yellow m a r r o w [1, 2, 26]; it m a y also demonstrate disease when bone scintigraphy findings are normal [2]. M R I has previously been used to assess the effects of treatment in patients with leukemia [l 3, 15-17, 23] and primary musculoskeletal tumors [6, 8, 11, 18, 24], but little is known a b o u t its ability to detect active bone m a r r o w neoplasia in patients who have undergone treatment for disseminated disease. Direct correlations between the findings provided by various spin echo and contrast-enhanced imaging techniques and by the histology of material obtained from a specific anatomic site are not generally available. In this study, we evaluated Tl-weighted, T zweighted, and short tau inversion recovery (STIR) image sequences, as well as static gadolinium diethylene triamine pentaacetic acid (Gd-DTPA)-enhanced, T1weighted images in 14 patients who had received chemotherapy, irradiation, or both for bone m a r r o w malignancies. We intended to determine the accuracy o f M R I in the diagnosis of active bone m a r r o w neoplasia in patients at risk for dissemination of primary disease.
Patients and methods We performed 21 MRI examinations in 14 patients with primary disseminated or secondary bone marrow malignancies. Six patients had suspected marrow metastases from neuroblastoma and three from Ewing's sarcoma; one patient had probable osteosarcomatous skip lesions. Three other patients with lymphoma and one with leukemia were also examined because of suspected bone marrow involvement. In all patients, bone marrow abnormalities were suspected because of MRI evidence of focal areas of marrow signal abnormalities in the form of hypointense regions on Tl-weighted images, high intensity on T2-weighted and STIR images, and enhancement of the lesions on Tl-weighted images obtained after administration of Gd-DTPA. Seven MRI studies were specified by protocol, and 14 were performed to monitor other disease dis9 1991 International Skeletal Society
80
S.L. Hanna et al. : MRI of bone marrow malignancy
Table 1. Clinical features of patients studied with magnetic resonance imaging (MRI) Patient Primary no. diagnosis 1 2 3 4 5 6 7 8 9 t0 11 12 t3 14
Neuroblastoma Neuroblastoma Neuroblastoma
Treatment
C, ABMT C, MIBG C, ABMT
Sites
Ilium Tibia Ilium, vertebra Neuroblastoma C, MIBG Ilium Neuroblastoma C, ABMT, Mandible MIBG Neuroblastoma C, ABMT Ilium Lymphoma C Ilium Lymphoma C Ilium Lymphoma C Ilium, femur Ewing's sarcoma C, RT Ilium Ewing's sarcoma C, RT Ilium Ewing's sarcoma C, RT Tibia Leukemia C Femur Osteosarcoma C Femur
Procedure (n) Biopsy (1) Biopsy (2) Biopsy (2), aspiration (1) Biopsy (1) Biopsy (t) Biopsy (1) Biopsy (2) Biopsy (2) Biopsy (2), excision (1) Biopsy (1) Biopsy (1) Biopsy (1) Biopsy (1) Amputation (l)
C, chemotherapy; ABMT, autologous bone marrow transplantation; MIBG, iodine 131 metaiodo benzylguanidine therapy; RT, external beam radiation therapy
covered on previous technetium 99m bone scintigrams and/or MRI studies. All patients had received chemotherapy; the four with neuroblastoma had undergone autologous bone marrow transplantation, and three had been treated with iodine 131 metaiodo benzylguanidine (131I-mIBG). The three patients with Ewing's sarcoma had received external beam radiation therapy to the primary site. The diagnoses, involved sites, diagnostic procedures, and details of therapy are shown in Table 1. Eighteen of the 21 MRI examinations were performed no more than 2 weeks before biopsy or surgical resection. One study was performed 4 weeks before a follow-up biopsy, but no further treatment was administered during that interval. Two MRI examinations were performed 4 and 10 days after needle biopsy. In one patient, MRI and marrow biopsy were performed before as well as 6 months after initiating chemotherapy; four imaging and histologic examinations were performed during the course of chemotherapy, eight others within 24 weeks after the completion of chemotherapy, and five within 7 weeks of the completion of combined chemotherapy and irradiation. In one patient, these procedures were performed 2.5 years and, in another, 10 years after chemotherapy and irradiation. Tissue for histologic examination was obtained from 21 anatomic sites, identified on the basis of focal bone marrow abnormalities apparent on MRI. Thirteen specimens were from an iliac crest, three from the femur, three from the tibia, one from the lumbar vertebral body, and one from the mandible. Eighteen of the tissue samples were collected by percutaneous needle biopsy (seven of which were fluoroscopically guided to sites delineated by MR signal abnormalities) using an 11- or 13-gauge Jamshidi needle; needle aspiration was used for one site. In addition, skip lesions corresponding to abnormal MR signals from the amputated femur of one patient with osteosarcoma and tissue from one excisional biopsy of the femur of a patient with lymphoma were examined microscopically. All biopsies and specimens contained marrow and were considered adequate for histologic interpretation. The MR studies were performed on a Siemens (Iselin, N J) system operating at 1.0 T. Tl-weighted sequences (TE 15, TR 600
ms) were used in 21 studies and T2-weighted sequences (TE 90, TR 2500 ms) in 14. STIR sequences (TE 22, TI 150, TR 2500 ms), which nullify the fat signal and increase the contrast of malignant tissue [4], were performed in 18 examinations. In 13 studies, the Tl-weighted sequences were repeated approximately 5 minutes after intravenous injection of 0.2 ml Gd-DTPA (Magnevist, Berlex) (0.1 mmol) per kg of body weight. Focal changes in the bone marrow signal on any or all of the imaging sequences were considered abnormal. Fourteen patients underwent 19 technetium 99m methylene diphosphonate (MDP) bone scintigrams. Three scintigrams were performed within 5 weeks and the other 16 within 3 weeks of MRI. Sixteen scans preceded biopsy or resection procedures by 1-4 weeks. The other three scintigrams were performed 1-4 weeks after biopsy. All patients also had roentgenographic and/or CT examinations including the biopsy sites. The MRI studies were evaluated independently from and without knowledge of the results of bone scintigrams, roentgenograms, or CT studies. Results Results o f histologic e x a m i n a t i o n s , M R I , a n d b o n e scan findings are p r e s e n t e d in Table 2. A l l M R I studies s h o w e d focal a b n o r m a l i t i e s o f b o n e m a r r o w signals o n one o r m o r e i m a g e sequences. O n T x - w e i g h t e d images, the a p p a r e n t a b n o r m a l i t y a p p e a r e d as a d a r k r e g i o n o f r e d u c e d signal in c o n t r a s t to fat's high signal. Ten o f the 14 T 2 - w e i g h t e d i m a g e s were a b n o r m a l , s h o w i n g c o n t r a s t reversal b e t w e e n the lesion a n d n o r m a l m a r r o w , in w h i c h the lesion a p p e a r e d relatively bright. B o n e m a r r o w a b n o r m a l i t i e s were r e c o g n i z e d in 17 o f the t 8 S T I R i m a g e s as r e g i o n s o f h i g h signal i n t e n s i t y in c o n t r a s t to the a b s e n t signal o f the m a r r o w fat. I n 10 o f the 13 lesions s t u d i e d b y G d - D T P A - e n h a n c e d , T l - w e i g h t e d , spin echo images, there was a n increase in the intensity o f the lesions as c o m p a r e d w i t h the n o n e n h a n c e d , T1w e i g h t e d images. T h i r t e e n o f the 19 r a d i o n u c l i d e b o n e scans also s h o w e d i n c r e a s e d u p t a k e at the sites o f susp e c t e d disease. L y t i c r o e n t g e n o g r a p h i c a n d / o r C T changes o f the c o r t e x o f the i n v o l v e d b o n e were a p p a r e n t in 10 patients. O n l y one child h a d a s s o c i a t e d soft-tissue involvement. O n h i s t o l o g i c e x a m i n a t i o n , 7 specimens, i n c l u d i n g the 1 o b t a i n e d b y needle a s p i r a t i o n a n d the 2 o b t a i n e d b y b i o p s y p r i o r to M R I , c o n t a i n e d active t u m o r ; 14 d i d not. T h e T l - w e i g h t e d i m a g e s o f all 7 t u m o r - p o s i t i v e lesions were a b n o r m a l , as were 4 o f the 6 T 2 - w e i g h t e d images, all o f 6 S T I R sequences, a n d all o f 5 G d - D T P A e n h a n c e d , T l - w e i g h t e d studies. Six o f 7 99mTc-MDP b o n e s c i n t i g r a m s were also positive, F o r lesions w i t h o u t active t u m o r , a b n o r m a l M R I signals were a p p a r e n t o n all 14 T ~ - w e i g h t e d images, 6 o f 8 T 2 - w e i g h t e d images, 11 o f 12 S T I R images, a n d 5 o f 8 G d - D T P A - e n h a n c e d , T a - w e i g h t e d images. E i g h t o f the 13 99mTc-MDP b o n e scans were positive. I n seven o f the t u m o r - n e g a t i v e lesions, t u m o r e x a m i n a t i o n disclosed v a r i o u s degrees o f h y p o p l a s i a o f h e m a t o p o i e t i c elements w i t h a r a n g e o f 0 % - 5 0 % cellularity; two specim e n s s h o w e d n o r m a l cellularity. Two h i s t o l o g i c a l l y negative specimens s h o w e d fibrosis, one was necrotic, a n d a n o t h e r d e m o n s t r a t e d a c o m b i n a t i o n o f necrosis a n d fibrosis. N i n e o f these specimens also s h o w e d replacem e n t o f the n o r m a l lipid c y t o p l a s m o f yellow b o n e m a r -
S.L. Hanna et al. : MRI of bone marrow malignancy
81
Table 2. Bone marrow imaging and histologic findings Patient
MRI sequences
99mTc bone scan
Bone marrow histology
no.
Tl-weighted
T2-weighted
STIR
Gd-DTPA
1
+
ND
--
--
2a
b
+ +
+
+ +
+ +
+ +
Tumor Tumor
3a b c
+ + +
ND ND
+ + ND
ND ND +
+
Tumor Hypocellular" Tumor
4
+
ND
+
ND
5
+
ND
+
ND
6
+
--
+
--
+
Hypocellular a
7a b
+ +
+ +
+ ND
ND +
Hypocellular a Hypocellular a
8a b
+ +
+ ND
+ +
ND -
+ ND +
9a b c
+ + +
ND + +
+ + +
+ + +
10
+
+
+
11
+
+
ND
Hypocellular a
Normal Hypocellular a
Tumor Normal" Hypocellular a Tumor Fibrosis
+
+ + ND +
ND
+
Necrosis, fibrosis
Fibrosis a
12
+
+
+
+
+
Necrosis
13
+
--
+
ND
+
Aspergillus
14
+
+
+
+
+
Tumor
ND, not done; + abnormal M R signal or bone scan; - normal " Specimens showed myxoid connective tissue matrix and serous atrophy of marrow fat
Fig. 1. A Coronal Tx-weighted, B gadolinium diethylene triamine pentaacetic acid (GdDTPA)-enhanced, T~-weighted, C T2-weighted, and D short tau inversion recovery (STIR) images of the left lower leg of a patient (no. 2) with neuroblastoma i year after completion of chemotherapy and immediately after treatment with 13II-mIBG. Low intensity lesion on the unenhanced Txweighted image (arrow), which was later subjected to biopsy under fluoroscopic control, increased slightly in brightness on the T2-weighted image enhanced with Gd-DTPA, and showed increased intensity on STIR sequence. Histologic examination showed active neuroblastoma r o w f a t cells b y s e r o u s f l u i d a n d a w a t e r y , m y x o i d c o n n e c t i v e tissue m a t r i x . O n e b i o p s y s p e c i m e n , a l t h o u g h free o f t u m o r , c o n t a i n e d Aspergillus h y p h a e a n d o t h e r changes of necrotizing osteomyelitis. O f t h e e i g h t s t u d i e s in w h i c h all f o u r M R I s e q u e n c e s w e r e p e r f o r m e d , six w e r e c h a r a c t e r i z e d b y s i g n a l a b n o r -
r e a l i t i e s o n all s e q u e n c e s . In t h r e e o f these, t h e m a r r o w s a m p l e s s h o w e d h i s t o l o g i c e v i d e n c e o f t u m o r (Fig. 1). In the remaining three, the specimens were histologically free o f t u m o r , w i t h f i b r o s i s in o n e , f i b r o s i s a n d s e r o u s c h a n g e s (Fig. 2) in o n e , a n d f r a n k n e c r o s i s in t h e last. F o u r p a t i e n t s e a c h u n d e r w e n t t w o b i o p s i e s o f a single
82
S.L. Hanna et al. : MRI of bone marrow malignancy
Fig. 2A-E. Transverse images of the pelvis of patient no. 10 who was treated for Ewing's sarcoma of the right sacral ala 2 years earlier. The low intensity lesion of the posterior aspect of the ilium, shown on the Tl-weighted image (A) and enhanced with Gd-DTPA administration (B), was bright on Tz-weighted (C) and STIR (D) images. E The bone marrow specimen, obtained by needle biopsy, showed hypoplasia of hematopoietic elements. The formerly lipid cytoplasm of the fat cells has been replaced by serous material, and the marrow spaces contain watery myxoid connective tissue (H& E, x800) a n a t o m i c site. In two instances, b o t h biopsies failed to yield histologic evidence o f t u m o r a l t h o u g h at least three M R I sequences o f each site were abnormal. In one patient, in w h o m the initial negative biopsy became positive, two sequences were a b n o r m a l at the time o f each biopsy. Tl-weighted a n d S T I R images remained abnormal at the biopsy site o f the fourth patient, whose lesion
was positive at the first biopsy but was negative at the follow-up biopsy.
Discussion M R I has p r o v e d effective in detecting m a l i g n a n t b o n e m a r r o w disease, whether T1- and T2-weighted, spin echo
S.L. Hanna et al. : MRI of bone marrow malignancy techniques [1-3, 9, 17, 19, 21, 22, 25], STIR, or chemical shift imaging [4, 20] are used. Gd-DTPA-enhanced, T1weighted imaging for diagnosis of bone neoplasms is still in the early stages of evaluation [5, 6]. Our investigation shows a similar effectiveness but is unable to address sensitivity, since selection for histological examination was based on discovery of a bone marrow abnormality on MRI. O f equal importance, however, is the ability of M R I to determine the presence or absence of active disease during or after treatment, which in our patients included chemotherapy and irradiation as well as bone marrow transplantation. Although all M R I diagnoses of metastases and myelomatous involvement of marrow were histologically confirmed in a previous large study [2], and several M R I studies on the effects of therapy on primary bone sarcomas are available [6, 8, 11, 18, 24], little is known about the specificity of M R I of bone marrow in patients undergoing various types o f anticancer therapy. Discrepancies between findings on M R I and radionuclide scintigrams [7, 10] raise questions about the relative effectiveness of these methods under these circumstances. Since all four M R I procedures were not performed during each study, their relative accuracy cannot be precisely quantitated. However, results of our comparison of histologic findings with M R I signal abnormalities in specific anatomic locations showed an apparently low specificity for M R I in the diagnosis of active bone marrow disease. Abnormal M R I signals were present on at least one imaging sequence of all 14 lesions that were negative for tumor. Indeed, the 99mTc skeletal scintigrams were at least as capable as M R I of identifying nonmalignant lesions. The accuracy of the methods used to obtain the histologic material may be questioned. Since all o f the 14 histologically nonmalignant specimens were obtained by means o f needle biopsy, sampling errors may have occurred. This possibility was recognized in previous M R I studies of lymphoma [12, 21] in which biopsies were confined to the iliac crests. In our investigation, however, sampling errors were presumably minimized by carefully selecting biopsy sites to conform to the regions of M R I signal abnormality. All biopsies contained adequate samples of bone marrow, and most of the falsepositive results could be accounted for by the histologic findings of necrosis or an increase in water content [14] associated with watery, myxoid changes of the marrow interstitium and replacement o f the lipid cytoplasm of the bone marrow fat by serous material. To some extent, the lack o f specificity of the M R I findings in our patients reflects the variable results o f M R I studies of treatment-associated changes in primary tumors. In one study of the effect of chemotherapy on bone sarcomas, no correlation was found between changes in the T2-weighted signal intensity and the histopathology [8]. In another, high signal on T2-weighted images was associated with radiation therapy changes rather than with active disease [24]. A high Tz-weighted signal intensity of the marrow cavity has, however, also been observed in patients with Ewing's sarcomas that were responsive to chemotherapy alone and, in several
83 of these tumors, was associated with histologic findings of serous atrophy [11]. Serous atrophy was also responsible for abnormal medullary signal remote from the primary tumor in several patients undergoing chemotherapy for osteosarcoma [18]. Finally, both viable tumor and vascularized, tumor-free areas of granulation tissue showed high signal intensity on Gd-DTPA-enhanced, Tl-weighted images of bone sarcomas following preoperative chemotherapy [6]. In a retrospective study of this size, in which not all M R I procedures could be performed on each patient, it is not possible to make statistically valid comparisons of their relative value. False-positive observations occurred with all of the M R I techniques employed as well as with the 99mTCbone scintigrams. Although bone scintigraphy led to fewer false-positive observations, M R I demonstrated one true-positive lesion that was not apparent on the radionuclide bone scan. In two patients with histologically positive lesions and in two others with negative lesions, T2-weighted images showed apparently normal bone marrow, whereas at least two other sequences demonstrated focal signal abnormalities. GdDTPA-enhanced images were normal in three patients without histologically demonstrated tumors; however, this technique gave false-positive results in more than 60% of the tumor-negative lesions for which it was employed. Therefore, although disseminated bone marrow neoplasia was readily detected by MRI, benign histologic findings associated with treatment could not be consistently differentiated from malignant disease with any of the imaging techniques employed. In this clinical setting, bone marrow abnormalities demonstrated on M R I studies should be interpreted with caution, and biopsy of the site in question may not always be appropriate.
Acknowledgements. This work was supported in part by National Cancer Institute grant PO1 CA-23099, Cancer Center Support (CORE) grant P30 CA-21765, and the American Lebanese Syrian Associated Charities (ALSAC).
References 1. Cohen MD, Klatte EC, Baehner R, et al. (1984) Magnetic resonance imaging of bone marrow disease in children. Radiology 151:715 2. Daffner RH, Lupetin AR, Dash N, et al. (1986) MRI in the detection of malignant infiltration of bone marrow. AJR 146:353 3. Dooms GC, Fisher MR, Hricak H, et al. (1985) Bone marrow imaging: magnetic resonance studies related to age and sex. Radiology t 55 :429 4. Dwyer AJ, Frank JA, Sank VJ, et al. (1988) Short-TI inversionrecovery pulse sequence: analysis and initial experience in cancer imaging. Radiology 168 : 827 5. Erlemann R, Reiser MF, Peters PE, et al. (1989) Musculoskeletal neoplasms: static and dynamic Gd-DTPA-enhanced MR imaging. Radiology 171 : 767 6. Erlemann R, Sciuk J, Bosse A, et al. (1990) Response of osteosarcoma and Ewing sarcoma to preoperative chemotherapy: assessment with dynamic and static MR imaging and skeletal scintigraphy. Radiology 175:791 7. Fletcher BD, Miraldi FD, Cheung N-KV (1989) Comparison
84 of radiolabeled monoclonal antibody and magnetic resonance imaging in the detection of metastatic neuroblastoma in bone marrow: preliminary results. Pediatr Radiol 20:72 8. Holscher HC, Bloem JL, Nooy MA, et al. (1990) The value of MR imaging in monitoring the effect of chemotherapy on bone sarcomas. AJR 154:763 9. Kangarloo H, Dietrich RB, Taira RT, et al. (1986) MR imaging of bone marrow in children. J Comput Assist Tomogr 10:205 10. Kattapuram SV, Khurana JS, Scott JA, E1-Khoury GY (1990) Negative scintigraphy with positive magnetic resonance imaging in bone metastases. Skeletal Radiol 19:113 11. Lemmi MA, Fletcher BD, Marina NM, et al. (1990) Use of MR imaging to assess results of chemotherapy for Ewing sarcoma. AJR 155:343 12. Linden A, Zankovich RE, Theissen P, Diehl V, Schicha H (1989) Malignant lymphoma: bone marrow imaging versus biopsy. Radiology 173:335 13. McKinstry CS, Steiner R, Young AT, et al. (1987) Bone marrow in leukemia and aplastic anemia: MRI imaging before, during and after treatment. Radiology 162:701 14. Mitchel DG, Burk DL Jr, Vinitski S, Rifkin MD (1987) The biophysical basis of tissue contrast in extracranial MR imaging. AJR 149:832 15. Moore SG, Gooding CA, Brasch RC, et al. (1986) Bone marrow in children with acute lymphocytic leukemia: MR relaxation times. Radiology 160:237 16. Nyman R, Rehn S, Glimelius B, et al. (1987) Magnetic resonance imaging in diffuse malignant bone marrow diseases. Acta Radiologica 28:199 17. Olson DO, Shields AF, Scheurich CJ, Porter BA, Moss AA
S i . Hanna et al. : MRI of bone marrow malignancy (1986) Magnetic resonance imaging of the bone marrow in patients with leukemia, aplastic anemia, and lymphoma. Invest Radio1 21 : 540 18. Pan G, Raymond AK, Carrasco CH, et al. (1990) Osteosarcoma: MR imaging after preoperative chemotherapy. Radiology 174:517 19. Porter BA, Shields AF, Olson DO (1986) Magnetic resonance imaging of bone marrow disorders. Radiol Clin North Am 24: 269 20. Rosen BR, Fleming DM, Kushner DC, et al. (1988) Hematologic bone marrow disorders: quantitative chemical shift MR imaging. Radiology 169: 799 21. Shields AF, Porter BA, Churchley S, et al. (1987) The detection of bone marrow involvement by lymphoma using magnetic resonance imaging. J Clin Oncol 5:225 22. Sugimura K, Yamasaki K, Kitagaki H, Tanaka Y, Kono M (1987) Bone marrow diseases of the spine: differentiation with T1 and T2 relaxation times in MR imaging. Radiology 165:541 23. Thomsen C, Sorensen PG, Karle H, Christoffersen P, Henriksen O (1987) Prolonged bone marrow Tl-relaxation in acute leukaemia. In vivo tissue characterization by magnetic resonance imaging. Magnetic Resonance Imaging 5:251 24. Vanel D, Lacombe M-J, Couanet D, et al. (1987) Musculoskeletal tumors: follow-up with MR imaging after treatment with surgery and radiation therapy. Radiology 164: 243 25. Yoshida H, Asai S, Yashiro N, Iio M (1985) MRI of bone marrow. Radiat Med 3 :47 26. Zimmer WD, Berquist TH, McLeod RA, et al. (1985) Bone tumors: magnetic resonance imaging versus computed tomography. Radiology 155 : 709
Skeletal Radiology
Magnetic resonance imaging of disseminated bone marrow disease in patients treated for malignancy S.L. Hanna, M.D., 1 B.D. Fletcher, M.D., 1 D.L. Fairclough, Dr. p.j.,2 J.H. Jenkins III, M.D., 3 A.H. Le, B.S. 1 Departments of 1Diagnostic Imaging, 2Biostatistics and Information Systems, and 3Pathology and Laboratory Medicine, St. Jude Children's Research Hospital and The University of Tennessee, Memphis, Tennessee, USA
Abstract. Magnetic resonance imaging ( M R I ) is a sensitive method for the diagnosis of bone m a r r o w abnormalities, but its usefulness in detecting active disseminated cancer in this tissue in treated patients has not been determined. We therefore examined 14 children who had been treated for disseminated bone m a r r o w involvement by neuroblastoma ( n = 6 ) , l y m p h o m a ( n = 3 ) , Ewing's sarcoma (n = 3), osteosarcoma (n = 1), and leukemia (n = 1). M R I studies were performed at 21 m a r r o w sites to evaluate residual or recurrent t u m o r and were correlated with histologic material from the same site. T1- and T2weighted sequences were employed in 21 and 14 studies, respectively; short tau inversion recovery (STIR) in 18; and static gadolinium diethylene triamine pentaacetic acid (Gd-DPTA)-enhanced, T~-weighted sequences in 13. All M R I studies showed an altered bone m a r r o w signal. Technetium 99m methylene diphosphonate (99mTc-MDP) bone scintigraphy was also performed (19 studies). On histologic examination, 7 m a r r o w specimens contained tumor, and 14 did not. O f the 7 tumor-positive lesions, all Ta-weighted, 4 of 6 T/-weighted, and all 6 S T I R sequences showed a b n o r m a l signal; all 5 GdDTPA-enhanced, Tl-weighted sequences showed enhancement of the lesion. However, abnormal signals were also observed on all Tx-weighted, 6 of 8 Tzweighted, 11 of 12 STIR, and 5 of 8 G d - D T P A - e n hanced, T~-weighted images of the tumor-negative sites. In this clinical setting, M R I did not consistently differentiate changes associated with treatment from malignant disease. Key words: Bone marrow, magnetic resonance studies - Bone marrow, biopsy - Magnetic resonance c o m p a r a tive studies Magnetic resonance in infants and children - Neoplasms
Address reprint requests to. Barry D. Fletcher, M.D., Department
of Diagnostic Imaging, St. Jude Children's Research Hospital, P.O. Box 318, Memphis, TN 38101, USA
Magnetic resonance imaging ( M R I ) is a valuable method for demonstrating malignant bone m a r r o w infiltration because of the excellent contrast between abnormal tissue and normal yellow m a r r o w [1, 2, 26]; it m a y also demonstrate disease when bone scintigraphy findings are normal [2]. M R I has previously been used to assess the effects of treatment in patients with leukemia [l 3, 15-17, 23] and primary musculoskeletal tumors [6, 8, 11, 18, 24], but little is known a b o u t its ability to detect active bone m a r r o w neoplasia in patients who have undergone treatment for disseminated disease. Direct correlations between the findings provided by various spin echo and contrast-enhanced imaging techniques and by the histology of material obtained from a specific anatomic site are not generally available. In this study, we evaluated Tl-weighted, T zweighted, and short tau inversion recovery (STIR) image sequences, as well as static gadolinium diethylene triamine pentaacetic acid (Gd-DTPA)-enhanced, T1weighted images in 14 patients who had received chemotherapy, irradiation, or both for bone m a r r o w malignancies. We intended to determine the accuracy o f M R I in the diagnosis of active bone m a r r o w neoplasia in patients at risk for dissemination of primary disease.
Patients and methods We performed 21 MRI examinations in 14 patients with primary disseminated or secondary bone marrow malignancies. Six patients had suspected marrow metastases from neuroblastoma and three from Ewing's sarcoma; one patient had probable osteosarcomatous skip lesions. Three other patients with lymphoma and one with leukemia were also examined because of suspected bone marrow involvement. In all patients, bone marrow abnormalities were suspected because of MRI evidence of focal areas of marrow signal abnormalities in the form of hypointense regions on Tl-weighted images, high intensity on T2-weighted and STIR images, and enhancement of the lesions on Tl-weighted images obtained after administration of Gd-DTPA. Seven MRI studies were specified by protocol, and 14 were performed to monitor other disease dis9 1991 International Skeletal Society
80
S.L. Hanna et al. : MRI of bone marrow malignancy
Table 1. Clinical features of patients studied with magnetic resonance imaging (MRI) Patient Primary no. diagnosis 1 2 3 4 5 6 7 8 9 t0 11 12 t3 14
Neuroblastoma Neuroblastoma Neuroblastoma
Treatment
C, ABMT C, MIBG C, ABMT
Sites
Ilium Tibia Ilium, vertebra Neuroblastoma C, MIBG Ilium Neuroblastoma C, ABMT, Mandible MIBG Neuroblastoma C, ABMT Ilium Lymphoma C Ilium Lymphoma C Ilium Lymphoma C Ilium, femur Ewing's sarcoma C, RT Ilium Ewing's sarcoma C, RT Ilium Ewing's sarcoma C, RT Tibia Leukemia C Femur Osteosarcoma C Femur
Procedure (n) Biopsy (1) Biopsy (2) Biopsy (2), aspiration (1) Biopsy (1) Biopsy (t) Biopsy (1) Biopsy (2) Biopsy (2) Biopsy (2), excision (1) Biopsy (1) Biopsy (1) Biopsy (1) Biopsy (1) Amputation (l)
C, chemotherapy; ABMT, autologous bone marrow transplantation; MIBG, iodine 131 metaiodo benzylguanidine therapy; RT, external beam radiation therapy
covered on previous technetium 99m bone scintigrams and/or MRI studies. All patients had received chemotherapy; the four with neuroblastoma had undergone autologous bone marrow transplantation, and three had been treated with iodine 131 metaiodo benzylguanidine (131I-mIBG). The three patients with Ewing's sarcoma had received external beam radiation therapy to the primary site. The diagnoses, involved sites, diagnostic procedures, and details of therapy are shown in Table 1. Eighteen of the 21 MRI examinations were performed no more than 2 weeks before biopsy or surgical resection. One study was performed 4 weeks before a follow-up biopsy, but no further treatment was administered during that interval. Two MRI examinations were performed 4 and 10 days after needle biopsy. In one patient, MRI and marrow biopsy were performed before as well as 6 months after initiating chemotherapy; four imaging and histologic examinations were performed during the course of chemotherapy, eight others within 24 weeks after the completion of chemotherapy, and five within 7 weeks of the completion of combined chemotherapy and irradiation. In one patient, these procedures were performed 2.5 years and, in another, 10 years after chemotherapy and irradiation. Tissue for histologic examination was obtained from 21 anatomic sites, identified on the basis of focal bone marrow abnormalities apparent on MRI. Thirteen specimens were from an iliac crest, three from the femur, three from the tibia, one from the lumbar vertebral body, and one from the mandible. Eighteen of the tissue samples were collected by percutaneous needle biopsy (seven of which were fluoroscopically guided to sites delineated by MR signal abnormalities) using an 11- or 13-gauge Jamshidi needle; needle aspiration was used for one site. In addition, skip lesions corresponding to abnormal MR signals from the amputated femur of one patient with osteosarcoma and tissue from one excisional biopsy of the femur of a patient with lymphoma were examined microscopically. All biopsies and specimens contained marrow and were considered adequate for histologic interpretation. The MR studies were performed on a Siemens (Iselin, N J) system operating at 1.0 T. Tl-weighted sequences (TE 15, TR 600
ms) were used in 21 studies and T2-weighted sequences (TE 90, TR 2500 ms) in 14. STIR sequences (TE 22, TI 150, TR 2500 ms), which nullify the fat signal and increase the contrast of malignant tissue [4], were performed in 18 examinations. In 13 studies, the Tl-weighted sequences were repeated approximately 5 minutes after intravenous injection of 0.2 ml Gd-DTPA (Magnevist, Berlex) (0.1 mmol) per kg of body weight. Focal changes in the bone marrow signal on any or all of the imaging sequences were considered abnormal. Fourteen patients underwent 19 technetium 99m methylene diphosphonate (MDP) bone scintigrams. Three scintigrams were performed within 5 weeks and the other 16 within 3 weeks of MRI. Sixteen scans preceded biopsy or resection procedures by 1-4 weeks. The other three scintigrams were performed 1-4 weeks after biopsy. All patients also had roentgenographic and/or CT examinations including the biopsy sites. The MRI studies were evaluated independently from and without knowledge of the results of bone scintigrams, roentgenograms, or CT studies. Results Results o f histologic e x a m i n a t i o n s , M R I , a n d b o n e scan findings are p r e s e n t e d in Table 2. A l l M R I studies s h o w e d focal a b n o r m a l i t i e s o f b o n e m a r r o w signals o n one o r m o r e i m a g e sequences. O n T x - w e i g h t e d images, the a p p a r e n t a b n o r m a l i t y a p p e a r e d as a d a r k r e g i o n o f r e d u c e d signal in c o n t r a s t to fat's high signal. Ten o f the 14 T 2 - w e i g h t e d i m a g e s were a b n o r m a l , s h o w i n g c o n t r a s t reversal b e t w e e n the lesion a n d n o r m a l m a r r o w , in w h i c h the lesion a p p e a r e d relatively bright. B o n e m a r r o w a b n o r m a l i t i e s were r e c o g n i z e d in 17 o f the t 8 S T I R i m a g e s as r e g i o n s o f h i g h signal i n t e n s i t y in c o n t r a s t to the a b s e n t signal o f the m a r r o w fat. I n 10 o f the 13 lesions s t u d i e d b y G d - D T P A - e n h a n c e d , T l - w e i g h t e d , spin echo images, there was a n increase in the intensity o f the lesions as c o m p a r e d w i t h the n o n e n h a n c e d , T1w e i g h t e d images. T h i r t e e n o f the 19 r a d i o n u c l i d e b o n e scans also s h o w e d i n c r e a s e d u p t a k e at the sites o f susp e c t e d disease. L y t i c r o e n t g e n o g r a p h i c a n d / o r C T changes o f the c o r t e x o f the i n v o l v e d b o n e were a p p a r e n t in 10 patients. O n l y one child h a d a s s o c i a t e d soft-tissue involvement. O n h i s t o l o g i c e x a m i n a t i o n , 7 specimens, i n c l u d i n g the 1 o b t a i n e d b y needle a s p i r a t i o n a n d the 2 o b t a i n e d b y b i o p s y p r i o r to M R I , c o n t a i n e d active t u m o r ; 14 d i d not. T h e T l - w e i g h t e d i m a g e s o f all 7 t u m o r - p o s i t i v e lesions were a b n o r m a l , as were 4 o f the 6 T 2 - w e i g h t e d images, all o f 6 S T I R sequences, a n d all o f 5 G d - D T P A e n h a n c e d , T l - w e i g h t e d studies. Six o f 7 99mTc-MDP b o n e s c i n t i g r a m s were also positive, F o r lesions w i t h o u t active t u m o r , a b n o r m a l M R I signals were a p p a r e n t o n all 14 T ~ - w e i g h t e d images, 6 o f 8 T 2 - w e i g h t e d images, 11 o f 12 S T I R images, a n d 5 o f 8 G d - D T P A - e n h a n c e d , T a - w e i g h t e d images. E i g h t o f the 13 99mTc-MDP b o n e scans were positive. I n seven o f the t u m o r - n e g a t i v e lesions, t u m o r e x a m i n a t i o n disclosed v a r i o u s degrees o f h y p o p l a s i a o f h e m a t o p o i e t i c elements w i t h a r a n g e o f 0 % - 5 0 % cellularity; two specim e n s s h o w e d n o r m a l cellularity. Two h i s t o l o g i c a l l y negative specimens s h o w e d fibrosis, one was necrotic, a n d a n o t h e r d e m o n s t r a t e d a c o m b i n a t i o n o f necrosis a n d fibrosis. N i n e o f these specimens also s h o w e d replacem e n t o f the n o r m a l lipid c y t o p l a s m o f yellow b o n e m a r -
S.L. Hanna et al. : MRI of bone marrow malignancy
81
Table 2. Bone marrow imaging and histologic findings Patient
MRI sequences
99mTc bone scan
Bone marrow histology
no.
Tl-weighted
T2-weighted
STIR
Gd-DTPA
1
+
ND
--
--
2a
b
+ +
+
+ +
+ +
+ +
Tumor Tumor
3a b c
+ + +
ND ND
+ + ND
ND ND +
+
Tumor Hypocellular" Tumor
4
+
ND
+
ND
5
+
ND
+
ND
6
+
--
+
--
+
Hypocellular a
7a b
+ +
+ +
+ ND
ND +
Hypocellular a Hypocellular a
8a b
+ +
+ ND
+ +
ND -
+ ND +
9a b c
+ + +
ND + +
+ + +
+ + +
10
+
+
+
11
+
+
ND
Hypocellular a
Normal Hypocellular a
Tumor Normal" Hypocellular a Tumor Fibrosis
+
+ + ND +
ND
+
Necrosis, fibrosis
Fibrosis a
12
+
+
+
+
+
Necrosis
13
+
--
+
ND
+
Aspergillus
14
+
+
+
+
+
Tumor
ND, not done; + abnormal M R signal or bone scan; - normal " Specimens showed myxoid connective tissue matrix and serous atrophy of marrow fat
Fig. 1. A Coronal Tx-weighted, B gadolinium diethylene triamine pentaacetic acid (GdDTPA)-enhanced, T~-weighted, C T2-weighted, and D short tau inversion recovery (STIR) images of the left lower leg of a patient (no. 2) with neuroblastoma i year after completion of chemotherapy and immediately after treatment with 13II-mIBG. Low intensity lesion on the unenhanced Txweighted image (arrow), which was later subjected to biopsy under fluoroscopic control, increased slightly in brightness on the T2-weighted image enhanced with Gd-DTPA, and showed increased intensity on STIR sequence. Histologic examination showed active neuroblastoma r o w f a t cells b y s e r o u s f l u i d a n d a w a t e r y , m y x o i d c o n n e c t i v e tissue m a t r i x . O n e b i o p s y s p e c i m e n , a l t h o u g h free o f t u m o r , c o n t a i n e d Aspergillus h y p h a e a n d o t h e r changes of necrotizing osteomyelitis. O f t h e e i g h t s t u d i e s in w h i c h all f o u r M R I s e q u e n c e s w e r e p e r f o r m e d , six w e r e c h a r a c t e r i z e d b y s i g n a l a b n o r -
r e a l i t i e s o n all s e q u e n c e s . In t h r e e o f these, t h e m a r r o w s a m p l e s s h o w e d h i s t o l o g i c e v i d e n c e o f t u m o r (Fig. 1). In the remaining three, the specimens were histologically free o f t u m o r , w i t h f i b r o s i s in o n e , f i b r o s i s a n d s e r o u s c h a n g e s (Fig. 2) in o n e , a n d f r a n k n e c r o s i s in t h e last. F o u r p a t i e n t s e a c h u n d e r w e n t t w o b i o p s i e s o f a single
82
S.L. Hanna et al. : MRI of bone marrow malignancy
Fig. 2A-E. Transverse images of the pelvis of patient no. 10 who was treated for Ewing's sarcoma of the right sacral ala 2 years earlier. The low intensity lesion of the posterior aspect of the ilium, shown on the Tl-weighted image (A) and enhanced with Gd-DTPA administration (B), was bright on Tz-weighted (C) and STIR (D) images. E The bone marrow specimen, obtained by needle biopsy, showed hypoplasia of hematopoietic elements. The formerly lipid cytoplasm of the fat cells has been replaced by serous material, and the marrow spaces contain watery myxoid connective tissue (H& E, x800) a n a t o m i c site. In two instances, b o t h biopsies failed to yield histologic evidence o f t u m o r a l t h o u g h at least three M R I sequences o f each site were abnormal. In one patient, in w h o m the initial negative biopsy became positive, two sequences were a b n o r m a l at the time o f each biopsy. Tl-weighted a n d S T I R images remained abnormal at the biopsy site o f the fourth patient, whose lesion
was positive at the first biopsy but was negative at the follow-up biopsy.
Discussion M R I has p r o v e d effective in detecting m a l i g n a n t b o n e m a r r o w disease, whether T1- and T2-weighted, spin echo
S.L. Hanna et al. : MRI of bone marrow malignancy techniques [1-3, 9, 17, 19, 21, 22, 25], STIR, or chemical shift imaging [4, 20] are used. Gd-DTPA-enhanced, T1weighted imaging for diagnosis of bone neoplasms is still in the early stages of evaluation [5, 6]. Our investigation shows a similar effectiveness but is unable to address sensitivity, since selection for histological examination was based on discovery of a bone marrow abnormality on MRI. O f equal importance, however, is the ability of M R I to determine the presence or absence of active disease during or after treatment, which in our patients included chemotherapy and irradiation as well as bone marrow transplantation. Although all M R I diagnoses of metastases and myelomatous involvement of marrow were histologically confirmed in a previous large study [2], and several M R I studies on the effects of therapy on primary bone sarcomas are available [6, 8, 11, 18, 24], little is known about the specificity of M R I of bone marrow in patients undergoing various types o f anticancer therapy. Discrepancies between findings on M R I and radionuclide scintigrams [7, 10] raise questions about the relative effectiveness of these methods under these circumstances. Since all four M R I procedures were not performed during each study, their relative accuracy cannot be precisely quantitated. However, results of our comparison of histologic findings with M R I signal abnormalities in specific anatomic locations showed an apparently low specificity for M R I in the diagnosis of active bone marrow disease. Abnormal M R I signals were present on at least one imaging sequence of all 14 lesions that were negative for tumor. Indeed, the 99mTc skeletal scintigrams were at least as capable as M R I of identifying nonmalignant lesions. The accuracy of the methods used to obtain the histologic material may be questioned. Since all o f the 14 histologically nonmalignant specimens were obtained by means o f needle biopsy, sampling errors may have occurred. This possibility was recognized in previous M R I studies of lymphoma [12, 21] in which biopsies were confined to the iliac crests. In our investigation, however, sampling errors were presumably minimized by carefully selecting biopsy sites to conform to the regions of M R I signal abnormality. All biopsies contained adequate samples of bone marrow, and most of the falsepositive results could be accounted for by the histologic findings of necrosis or an increase in water content [14] associated with watery, myxoid changes of the marrow interstitium and replacement o f the lipid cytoplasm of the bone marrow fat by serous material. To some extent, the lack o f specificity of the M R I findings in our patients reflects the variable results o f M R I studies of treatment-associated changes in primary tumors. In one study of the effect of chemotherapy on bone sarcomas, no correlation was found between changes in the T2-weighted signal intensity and the histopathology [8]. In another, high signal on T2-weighted images was associated with radiation therapy changes rather than with active disease [24]. A high Tz-weighted signal intensity of the marrow cavity has, however, also been observed in patients with Ewing's sarcomas that were responsive to chemotherapy alone and, in several
83 of these tumors, was associated with histologic findings of serous atrophy [11]. Serous atrophy was also responsible for abnormal medullary signal remote from the primary tumor in several patients undergoing chemotherapy for osteosarcoma [18]. Finally, both viable tumor and vascularized, tumor-free areas of granulation tissue showed high signal intensity on Gd-DTPA-enhanced, Tl-weighted images of bone sarcomas following preoperative chemotherapy [6]. In a retrospective study of this size, in which not all M R I procedures could be performed on each patient, it is not possible to make statistically valid comparisons of their relative value. False-positive observations occurred with all of the M R I techniques employed as well as with the 99mTCbone scintigrams. Although bone scintigraphy led to fewer false-positive observations, M R I demonstrated one true-positive lesion that was not apparent on the radionuclide bone scan. In two patients with histologically positive lesions and in two others with negative lesions, T2-weighted images showed apparently normal bone marrow, whereas at least two other sequences demonstrated focal signal abnormalities. GdDTPA-enhanced images were normal in three patients without histologically demonstrated tumors; however, this technique gave false-positive results in more than 60% of the tumor-negative lesions for which it was employed. Therefore, although disseminated bone marrow neoplasia was readily detected by MRI, benign histologic findings associated with treatment could not be consistently differentiated from malignant disease with any of the imaging techniques employed. In this clinical setting, bone marrow abnormalities demonstrated on M R I studies should be interpreted with caution, and biopsy of the site in question may not always be appropriate.
Acknowledgements. This work was supported in part by National Cancer Institute grant PO1 CA-23099, Cancer Center Support (CORE) grant P30 CA-21765, and the American Lebanese Syrian Associated Charities (ALSAC).
References 1. Cohen MD, Klatte EC, Baehner R, et al. (1984) Magnetic resonance imaging of bone marrow disease in children. Radiology 151:715 2. Daffner RH, Lupetin AR, Dash N, et al. (1986) MRI in the detection of malignant infiltration of bone marrow. AJR 146:353 3. Dooms GC, Fisher MR, Hricak H, et al. (1985) Bone marrow imaging: magnetic resonance studies related to age and sex. Radiology t 55 :429 4. Dwyer AJ, Frank JA, Sank VJ, et al. (1988) Short-TI inversionrecovery pulse sequence: analysis and initial experience in cancer imaging. Radiology 168 : 827 5. Erlemann R, Reiser MF, Peters PE, et al. (1989) Musculoskeletal neoplasms: static and dynamic Gd-DTPA-enhanced MR imaging. Radiology 171 : 767 6. Erlemann R, Sciuk J, Bosse A, et al. (1990) Response of osteosarcoma and Ewing sarcoma to preoperative chemotherapy: assessment with dynamic and static MR imaging and skeletal scintigraphy. Radiology 175:791 7. Fletcher BD, Miraldi FD, Cheung N-KV (1989) Comparison
84 of radiolabeled monoclonal antibody and magnetic resonance imaging in the detection of metastatic neuroblastoma in bone marrow: preliminary results. Pediatr Radiol 20:72 8. Holscher HC, Bloem JL, Nooy MA, et al. (1990) The value of MR imaging in monitoring the effect of chemotherapy on bone sarcomas. AJR 154:763 9. Kangarloo H, Dietrich RB, Taira RT, et al. (1986) MR imaging of bone marrow in children. J Comput Assist Tomogr 10:205 10. Kattapuram SV, Khurana JS, Scott JA, E1-Khoury GY (1990) Negative scintigraphy with positive magnetic resonance imaging in bone metastases. Skeletal Radiol 19:113 11. Lemmi MA, Fletcher BD, Marina NM, et al. (1990) Use of MR imaging to assess results of chemotherapy for Ewing sarcoma. AJR 155:343 12. Linden A, Zankovich RE, Theissen P, Diehl V, Schicha H (1989) Malignant lymphoma: bone marrow imaging versus biopsy. Radiology 173:335 13. McKinstry CS, Steiner R, Young AT, et al. (1987) Bone marrow in leukemia and aplastic anemia: MRI imaging before, during and after treatment. Radiology 162:701 14. Mitchel DG, Burk DL Jr, Vinitski S, Rifkin MD (1987) The biophysical basis of tissue contrast in extracranial MR imaging. AJR 149:832 15. Moore SG, Gooding CA, Brasch RC, et al. (1986) Bone marrow in children with acute lymphocytic leukemia: MR relaxation times. Radiology 160:237 16. Nyman R, Rehn S, Glimelius B, et al. (1987) Magnetic resonance imaging in diffuse malignant bone marrow diseases. Acta Radiologica 28:199 17. Olson DO, Shields AF, Scheurich CJ, Porter BA, Moss AA
S i . Hanna et al. : MRI of bone marrow malignancy (1986) Magnetic resonance imaging of the bone marrow in patients with leukemia, aplastic anemia, and lymphoma. Invest Radio1 21 : 540 18. Pan G, Raymond AK, Carrasco CH, et al. (1990) Osteosarcoma: MR imaging after preoperative chemotherapy. Radiology 174:517 19. Porter BA, Shields AF, Olson DO (1986) Magnetic resonance imaging of bone marrow disorders. Radiol Clin North Am 24: 269 20. Rosen BR, Fleming DM, Kushner DC, et al. (1988) Hematologic bone marrow disorders: quantitative chemical shift MR imaging. Radiology 169: 799 21. Shields AF, Porter BA, Churchley S, et al. (1987) The detection of bone marrow involvement by lymphoma using magnetic resonance imaging. J Clin Oncol 5:225 22. Sugimura K, Yamasaki K, Kitagaki H, Tanaka Y, Kono M (1987) Bone marrow diseases of the spine: differentiation with T1 and T2 relaxation times in MR imaging. Radiology 165:541 23. Thomsen C, Sorensen PG, Karle H, Christoffersen P, Henriksen O (1987) Prolonged bone marrow Tl-relaxation in acute leukaemia. In vivo tissue characterization by magnetic resonance imaging. Magnetic Resonance Imaging 5:251 24. Vanel D, Lacombe M-J, Couanet D, et al. (1987) Musculoskeletal tumors: follow-up with MR imaging after treatment with surgery and radiation therapy. Radiology 164: 243 25. Yoshida H, Asai S, Yashiro N, Iio M (1985) MRI of bone marrow. Radiat Med 3 :47 26. Zimmer WD, Berquist TH, McLeod RA, et al. (1985) Bone tumors: magnetic resonance imaging versus computed tomography. Radiology 155 : 709
