Case Reports / Journal of Clinical Neuroscience 19 (2012) 1167–1169 1167

pontine and/or extrapontine myelinolysis) in 25 patients. J Neurol NeurosurgPsychiatry 2011;82:326–31.

2. Maraganore DM, Folger WN, Swanson JW, et al. Movement disorders as sequelaeof central pontine myelinolysis: report of three cases. Mov Disord 1992;7:142–8.

3. Post B, van Gool WA, Tijssen MA. Transient parkinsonism in isolatedextrapontine myelinolysis. J Neurol Sci 2009;30:325–8.

4. Kim JS, Lee KS, Han SR, et al. Decreased striatal dopamine transporter binding ina patient with extrapontine myelinolysis. Mov Disord 2003;18:342–5.

doi:10.1016/j.jocn.2011.08.043

⇑ Corresponding author. Tel.: +32 9 332 30 28; fax: +32 9 332 38 07.E-mail address: ingeborg.goethals@ugent.be (I. Goethals).

5. Nagamitsu S, Matsuishi T, Yamashita Y, et al. Extrapontine myelinolysis withparkinsonism after rapid correction of hyponatremia: high cerebrospinal fluidlevels of homovanillic acid and successful dopaminergic treatment. J NeuralTransm 1999;106:949–53.

6. Seah AB, Chan LL, Wong MC, et al. Evolving spectrum of movement disorders inextrapontine and central pontine myelinolysis. Parkinsonism Relat Disord2002;9:117–9.

Progressive multifocal leukoencephalopathy (PML) mimicking high-grade gliomaon delayed F-18 FDG PET imaging

Koen Mertens a, Marjan Acou b, Caroline Van den Broecke c, Roel Nuyts d, Dirk Van Roost d, Eric Achten b,Ingeborg Goethals a,⇑a Department of Nuclear Medicine, Polikliniek 7, Ghent University Hospital, De Pintelaan 185, Gent 9000, Belgiumb Department of Radiology, Ghent University Hospital, Gent, Belgiumc Department of Pathology, Ghent University Hospital, Gent, Belgiumd Department of Neurosurgery, Ghent University Hospital, Gent, Belgium

a r t i c l e i n f o

Article history:Received 8 August 2011Accepted 13 August 2011

Keywords:Delayed imagingF-18 fluorodeoxyglucosePositron emission tomographyProgressive multifocal leukoencephalopathy

a b s t r a c t

The purpose of our study was to determine the increase in F-18 fluorodeoxyglucose (FDG) uptake in apatient with progressive multifocal leukoencephalopathy (PML) between early and late scan times usingpositron emission tomography (PET) imaging with F-18 FDG at conventional (60 minutes [min] afterinjection, PET60) and delayed (300 min after injection, PET300) intervals. PET60 and PET300 imaging wasperformed on a pathologically proven PML lesion. The PML lesion in the posterior fossa exhibited anincrease in F-18 FDG uptake of 52% between early and late times, which was in the range of that inhigh-grade gliomas. Thus, dual-time-point PET with F-18 FDG may not be able to differentiate betweeninfectious and malignant brain lesions.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Progressive multifocal leukoencephalopathy (PML) is a progres-sive, demyelinating disease caused by the reactivation of the JohnCunningham (JC) virus in immunocompromised patients.1 Clinicalfeatures are non-specific and the diagnosis is often considered forthe first time based on MRI findings. White matter involvementusually appears on MRI as asymmetric multifocal lesions withoutcontrast enhancement. However, conventional MRI is not specificenough for the differential diagnosis of single lesions. Complemen-tary to conventional MRI, proton magnetic resonance (MR) spec-troscopy (MRS) and positron emission tomography (PET) imagingcan be used for metabolic characterisation. Definite diagnosis isbased on cerebrospinal fluid polymerase chain reaction for JC virusDNA and a brain biopsy.2 Here, we report a patient with PML pre-senting with the acute onset of focal neurological signs and a largenon-enhancing lesion in the posterior fossa on MRI. The lesion wasfurther characterised by dual-time-point F-18 fluorodeoxyglucose(FDG) PET imaging. To our knowledge this is the first reported caseof PML studied by F-18 FDG PET at delayed intervals.

2. Case report

The patient is a 52-year-old man with a history of heart trans-plantation. In November 2008, he developed diplopia and a leftcerebellar syndrome. Conventional MRI at the time of admissionrevealed hyperintense fluid-attenuated inversion recovery andhypointense T1-weighted signal changes without contrastenhancement in the left cerebellar hemisphere, extending in tothe middle cerebellar peduncle. On diffusion-weighted imaging(DWI), the lesion was hyperintense, with high signal on the corre-sponding apparent diffusion coefficient (ADC) map images, indicat-ing T2 shine through. The absence of diffusion restriction ruled outrecent ischemia. Because the patient was immunocompromised,the diagnosis of PML was considered. MRS was performed to differ-entiate the lesion from a low-grade glioma. On MRS, however,there was no significant increase in choline (Cho), which is presentin glial brain tumors. The metabolic profile showed a decrease inN-acetylaspartate (NAA) and high levels of lactate (Fig. 1).

The large lesion provided an opportunity to determine the met-abolic profile further, and we characterised the lesion with dual-time-point F-18 FDG PET studies. On the PET60 images, the mostprominent finding was a high uptake in the pons; however, on con-ventional MRI or DWI, no signal abnormalities were observed in thebrain stem. Analysis of the PET60 and PET300 images further indi-cated that the activity in the pons increased with time. The increasebetween early and late scan times was 38%. In addition, there was

an area of decreased uptake in the cerebellum on the PET60 images,which corresponded with the lesion in the left middle cerebellarpeduncle on MRI. At delayed intervals, uptake in the lesion washigher than in the normal surrounding cerebellum. The increasein uptake over time was 52% (Fig. 2). Following MRI and PETimaging, an open biopsy of the lesion in the left cerebellar pedunclewas performed. The pathological diagnosis of PML was made be-cause glial cells harboring JC viral inclusions were detected (Supple-mentary Fig. 1). In the weeks following diagnosis, the patient’sneurological condition worsened and he died soon afterwards.

3. Discussion

On MRI, typical PML lesions are diffuse, asymmetric and locatedin the supratentorial white matter.2 For atypical lesions, MRS canprovide additional information. The spectrum is characterized by amoderate increase of the Cho signal in relation to demyelination.The presence of high levels of lactate probably reflect necrosis. Thedecrease of NAA, a neuronal marker, can be related to neuronal dam-age or death in a necrotic area.3–5 There are only a few reports of F-18FDG PET in PML.6–8 On these PET images (at conventional intervals),

Fig. 1. (A–D) Axial (A) hyperintense fluid-attenuated inversion recovery and (B) hypointense T1-weighted signal changes without contrast enhancement in the left cerebellarhemisphere; (C) hyperintense diffusion-weighted imaging with (D) high signal on the corresponding apparent diffusion coefficient map in a patient with progressivemultifocal leukoencephalopathy. (E) Single voxel spectroscopy performed at short echo times (time to echo [TE] = 30 ms) and (F) a long TE (135 ms) showed a decrease in N-acetylaspartate (NAA), with a NAA/creatine ratio of 0.56, high levels of lactate and a moderate increase in choline (Cho), with a Cho/creatine ratio of 1.29 (TE = 135 ms).

1168 Case Reports / Journal of Clinical Neuroscience 19 (2012) 1167–1169

both high and low uptake of F-18 FDG in PML lesions has been re-ported.6–8 Our findings agree with those in the literature: on thePET60 images, the PML lesion in the pons was hypermetabolic,whereas the lesion in the middle cerebellar peduncle was hypomet-abolic. However, it is well known that delaying PET imaging well be-yond the standard imaging time better distinguishes metabolicallyabnormal from normal brain tissue.9 Additionally, dual-time-pointPET provides an opportunity to examine the increase in uptake be-tween early and late scan times. Using ratios of tumor standard up-take value (SUV) to reference region SUV, Spence and colleagueshave shown that the increase in uptake by gliomas was roughly20% between early and late scan times. In PML, we found an increasein uptake over time of 38% for the lesion in the pons, and 52% for thelesion in the middle cerebellar peduncle. Moreover, in comparisonwith our institution’s data obtained in low- and high-grade gliomas,the percentage increase for both PML lesions was in the range of thatfor high-grade gliomas. Therefore, infectious brain lesions may havea similar or higher increase in lesion-to-background activity withtime than high-grade gliomas. This may hamper the use of delayedF-18 FDG PET imaging to discriminate benign from malignantlesions.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.jocn.2011.08.043.

References

1. Tan CS, Koralnik IJ. Progressive multifocal leukoencephalopathy and otherdisorders caused by JC virus: clinical features and pathogenesis. Lancet Neurol2010;9:425–37.

2. Shah R, Bag A, Chapman P, et al. Imaging manifestations of progressivemultifocal leukoencephalopathy. Clin Radiol 2010;65:431–9.

3. Chang L, Ernst T, Tornatore C, et al. Metabolite abnormalities in progressivemultifocal leukoencephalopathy by proton magnetic resonance spectroscopy.Neurology 1997;48:836–45.

4. Simone IL, Federico F, Tortorella C, et al. Localised 1H–MR spectroscopy formetabolic characterisation of diffuse and focal brain lesions in patients infectedwith HIV. J Neurol Neurosurg Psychiatry 1998;64:516–23.

5. Salvan AM, Confort-Gouny S, Cozzone PJ, et al. Atlas of brain proton magneticresonance spectra. Part III: Viral infections. J Neuroradiol 1999;26:154–61.

6. Pierce MZ, Johnson MD, Maciunas RJ, et al. Evaluating contrast-enhancing brainlesions in patients with AIDS by using positron emission tomography. Ann InternMed 1995;123:594–8.

7. Heald AE, Hoffman JM, Bartlett JA, et al. Differentiation of central nervous systemlesions in AIDS patients using positron emission tomography (PET). Int J STD AIDS1996;7:337–46.

8. O’Doherty MJ, Barrington SF, Campbell M, et al. PET scanning and the humanimmunodeficiency virus-positive patient. J Nucl Med 1997;38:1575–83.

9. Spence AM, Muzi M, Mankoff DA, et al. 18F-FDG PET of gliomas at delayedintervals: improved distinction between tumor and normal gray matter. J NuclMed 2004;45:1653–9.

doi:10.1016/j.jocn.2011.08.043

Fig. 2. A graph of the percentage change in F-18 fluorodeoxyglucose (FDG) uptake between early and late scan times for gliomas and the progressive multifocalleukoencephalopathy (PML) lesions. Uptake is expressed as the ratio of late to early tumor counts (both normalised to a reference region). The increase in uptake was 52% forthe lesion in the left middle cerebellar peduncle, and 38% for the lesion in the pons. (A) Uptake values were in the range of those seen in high-grade gliomas. (B) Axial positronemission tomography (PET) image at 60 minutes (PET60) showing increased uptake in the pons and (C) decreased uptake in the left cerebellar hemisphere. At delayedintervals, uptake is high in both lesions compared to the surrounding normal tissue.

Case Reports / Journal of Clinical Neuroscience 19 (2012) 1167–1169 1169