CASE REPORT

Spinal intramedullary metastasis of medulloblastomaat initial diagnosis

Tomoo Inoue & Toshihiro Kumabe &

Toshiyuki Takahashi & Takeshi Nakajima &

Mika Watanabe & Teiji Tominaga

Received: 21 August 2005 / Published online: 28 October 2006# Springer-Verlag 2006

AbstractCase report Spinal magnetic resonance imaging of a 4-year-old boy with medulloblastoma at the initial presenta-tion showed intramedullary lesion without enhancementeffect and slight cord swelling from C-5 to T-1. Aftercomplete response to the initial therapy, this lesion recurredand slowly expanded. Cervical 11C-methionine-positronemission tomography could establish the diagnosis ofintramedullary metastasis.Conclusion Spinal intramedullary metastasis of medullo-blastoma at initial diagnosis is extremely rare, but must beconsidered.

Keywords Dissemination . Intramedullary metastasis .

Magnetic resonance imaging .Medulloblastoma .

Methionine positron emission tomography . Initial diagnosis

Introduction

Medulloblastoma is known to invade the neuraxis, but spinalintramedullary metastasis is rare. In this study, we describe acase of medulloblastoma disseminated into the intramedul-lary cervical spinal cord at the initial presentation.

Case report

A 4-year-old boy was first admitted in 1999 because ofheadache and snoring. Magnetic resonance (MR) imagingdemonstrated an irregular enhanced tumor almost filling thefourth ventricle and marked obstructive hydrocephalus(Fig. 1). Spinal T1-weighted MR imaging with gadoliniumshowed an intramedullary hypointensity lesion and slightspinal cord swelling from C-5 to T-1 (Fig. 2a), although thisfinding was not initially recognized as a metastasis. Thefourth ventricle tumor was removed completely, and thehistological diagnosis was medulloblastoma. Postoperative-ly, he received three courses of chemotherapy with cisplatinand etoposide followed by irradiation consisting of 24 Gyto the whole brain and whole spine, and 30 Gy to theposterior fossa. Three more courses of the same chemo-therapy were then given. MR imaging showed the tumor inthe cervical spinal cord had disappeared in June 2000(Fig. 2b), but demonstrated regrowth of the spinal lesion inJuly 2001 (Fig. 2c). The lesion slowly expanded, but notreatment was performed because he showed no neurolog-ical deterioration. 11C-methionine-positron emission tomog-raphy (MET-PET) indicated the mean standardized uptakevalue (SUV) was 1.62 in the spinal lesion in July 2003.SUV, which indicates the tissue activity of 11C-methionineper unit body weight, was calculated as (average concen-tration of radioactivity in a range of interest)/(injectedradioisotope activity/body weight) × calibration factor(Fig. 3a). The lesion gradually expanded, but generalphysical examination was unremarkable until December2003, when he suffered from acute posterior neck painwhen playing football with no signs of gait disturbance, orbowel and bladder dysfunction. MR imaging showednotable spinal cord swelling from C-4 to T-4 (Fig. 2d)without any other tumor recurrence including the fourth

Childs Nerv Syst (2007) 23:113–116DOI 10.1007/s00381-006-0167-7

T. Inoue : T. Kumabe (*) : T. Takahashi : T. Nakajima :T. TominagaDepartment of Neurosurgery,Tohoku University Graduate School of Medicine,1-1, Seiryo-machi, Aoba-ku,Sendai 980-8574, Japane-mail: kuma@nsg.med.tohoku.ac.jp

M. WatanabeDepartments of Pathology,Tohoku University Graduate School of Medicine,Sendai, Japan

ventricle. MET-PET disclosed the same area had a meanSUV of 2.16 and a maximum SUV of 3.29 (Fig. 3b). Wethought the lesion was not radiation necrosis but tumorprogression because the value of SUV was increasedcompared to the last time. Therefore, C5-6 osteoplasticlaminotomy and biopsy of the cervical spinal lesion wereperformed. The surface of spinal cord was intact macro-scopically, and we carried out a tumor biopsy via theposterior median fissure. The surgical specimen was

consistent with disseminated medulloblastoma (Fig. 4). Hecomplained of momentary lower limb disturbance, andsubsequently received three courses of chemotherapy usingifosfamide, cisplatin, and etoposide. The spinal cord wasshrinking after the treatment, but MR imaging in February2005 demonstrated augmentation of the lesion. He died oftumor dissemination throughout all parts of the centralnervous system in June 2005.

Discussion

Medulloblastoma often initially showed leptomeningealdissemination [1], although the possibility of metastasisinto the spinal cord at diagnosis was not mentioned as far aswe know. Only two similar cases of intramedullary spinallesion histologically confirmed as medulloblastoma havebeen reported [2, 3]. Unlike the previous cases, the spinalintramedullary lesion could be identified retrospectively atthe initial presentation. If we had considered the first MRfinding as metastasis, the patient could have been receivedlocal radiation boost in the initial treatment, or additionalchemotherapy and radiation therapy could have beenperformed if the reappeared spinal lesion was diagnosedproperly as recurrence in 2001. Assuming that the treatmentwas done, his terminal stage might have been prolonged.

Fig. 1 Axial T1-weighted magnetic resonance (MR) image withgadolinium-diethylenetriaminepenta-acetic acid (Gd-DTPA) on admis-sion, demonstrating a fourth ventricle tumor

Fig. 2 Sagittal T1-weighted MR image with Gd-DTPA of the spine.Before the initial treatment, a hypointense lesion (arrows) and slightspinal cord swelling from C-5 to T-1 was depicted (a). The lesiondisappeared in June 2000 (b). Follow-up MR image revealed regrowth

of the spinal lesion (arrowheads) in July 2001 (c). MR image obtainedjust before the open biopsy for the cervical spinal lesion in December2003, demonstrated a massive intramedullary spinal cord tumorextending from C-4 to T-4 with scattered enhancement (d)

114 Childs Nerv Syst (2007) 23:113–116

Though the region of gadolinium enhancement isconsidered to be consistently well to the tumor mass,magnetic resonance imaging (MRI) cannot evaluate themetabolism of a lesion. PET is possible to estimate themetabolism of brain tumors in vivo by variety of positrontracers [4]. Though it remained unclear, the mechanism ofaccumulation of MET in tissues are thought to presentdisruption of the blood brain barrier, abundant vessels, and

active carrier-mediated transport across the cell membrane[5]. Therefore, its accumulation is thought to increaseprotein synthesis by neoplastic tissue proliferation. Asignificant uptake of MET-PET is considered to reflect theexistent of tumor cells. MET-PET, an imaging techniqueproviding metabolic data, has been reported to play animportant role in improving diagnostic procedures fordifferentiating malignant cells from radiation necrosis.Tsuyuguchi et al. [6, 7] reported the availability of MET-PET for differentiating between recurrent malignant braintumor and radiation necrosis. According to their report, themean SUV is about 2.5 in the first group and about 1.8 inthe second group. In detecting tumor recurrence, theysuggested MET-PET as the high sensitivity and accuracy,and no false–negative cases were observed. In falsepositive cases, glial fibrillary acid protein immunostainingas the marker of reactive gliosis showed positive findingsaround the necrotic tissues. It revealed that proliferativechange in glial cells might be a mechanism of MET uptake[7]. In contrast to brain tumors, there are few reports ofPET findings about spinal cord tumors due to its lowcontrast result with in the small space of vertebral bodies.[8, 9] In our case, cervical MET-PET successfully identifieda metabolically active lesion and established the diagnosisof spinal intramedullary metastasis. The present caseindicates that the combination of spinal MR imagingand MET-PET allows comprehensive evaluation of spinalneoplasms.

In conclusion, spinal intramedullary metastasis of me-dulloblastoma at initial diagnosis is extremely rare, butmust be considered.

Fig. 4 Photomicrograph of the surgical specimen from thecervical spinal lesion, showing typical findings of medulloblas-toma as a highly cellular tumor with hyperchromatic nuclei,scant cytoplasm, and a Homer–Wright rosette in the center. (Hand E, original magnification ×100)

Fig. 3 11C-methionine-positronemission tomography scans inJuly (a) and in December (b)2003. The second examinationrevealed obvious uptake of thetracer in the lesion

Childs Nerv Syst (2007) 23:113–116 115

References

1. Meyers SP, Wildenhain SL, Chang JK, Bourekas EC, Beattie PF,Korones DN, Davis D, Pollack IF, Zimmerman RA (2000)Postoperative evaluation for disseminated medulloblastoma involv-ing the spine: contrast-enhanced MR findings, CSF cytologicanalysis, timing of disease occurrence, and patient outcomes. AmJ Neuroradiol 21: 1757–1765

2. Barnwell SL, Edwards MS (1986) Spinal intramedullary spread ofmedulloblastoma. J Neurosurg 65:253–255

3. Zumpano BJ (1978) Spinal intramedullary metastatic medulloblas-toma. Case report. J Neurosurg 48:632–635

4. Patronas NJ, Di Chiro G, Kufta C, Bairamian D, Kornblith PL,Simon R, Larson SM (1985) Prediction of survival in gliomapatients by means of positron emission tomography. J Neurosurg62:816–822

5. Kubota K, Matsuzawa T, Fujiwara T, Abe Y, Hatazawa J, KawabeH, Miyazawa H, Sato T, Ido T, Ishiwata K (1985) An experimentalstudy on differential diagnosis of cancer from inflammation using11C-L-methionine. Kaku Igaku 22:1327–1332

6. Tsuyuguchi N, Sunada I, Iwai I, Yamanaka K, Tanaka K,Takami T, Otsuka Y, Sakamoto S, Ohata K, Goto T, Hara M(2003) Methionine positron emission tomography of recurrentmetastatic brain tumor and radiation necrosis after stereotacticradiology: is a differential diagnosis possible? J Neurosurg98:1056–1064

7. Tsuyuguchi N, Takami T, Sunada I, Iwai Y, Yamanaka K, TanakaK, Nishikawa M, Ohata K, Torii K, Morino M, Nishio A, Hara M(2004) Methionine positron emission tomography for differentia-tion of recurrent brain tumor and radiation necrosis afterstereotactic radiosurgery in malignant glioma. Ann Nucl Med18:291–296

8. Higano S, Shishido F, Nagashima M, Tomura N, Murakami M,Inugami A, Fujita H, Tabata K, Yasui N, Uemura K (1990) PETevaluation of spinal cord tumor using 11C-methionine. J ComputAssist Tomogr 14:297–299

9. Sasajima T, Mineura K, Itoh Y, Kowada M, Hatazawa J, Ogawa T,Uemura K (1996) Spinal cord ependymoma: a positron emissiontomographic study with (11C-methyl)-L-methionine.Neuroradiology 38:53–55

116 Childs Nerv Syst (2007) 23:113–116