Brain MRI findings in aspartylglucosaminuria

+Model NEURAD-515; No. of Pages 13

ARTICLE IN PRESS

Journal of Neuroradiology (2015) xxx, xxx—xxx

Available online at

ScienceDirect www.sciencedirect.com

ORIGINAL ARTICLE

Brain MRI findings in aspartylglucosaminuria Anna M. Tokola a,∗, Laura E. Åberg b, Taina H. Autti a a

HUS Medical Imaging Center, Radiology, University of Helsinki and Helsinki University Hospital, P.O. Box 340, FI-00029 HUS, Helsinki, Finland b Department of Psychiatry, Jorvi Hospital, P.O. Box 827, FI-00029 HUS, Finland

KEYWORDS Aspartylglucosaminuria; Lysosomal storage disorder; Inborn errors of metabolism; Magnetic resonance imaging; Intellectual disability

Summary Background and purpose: The aim of this study was to identify characteristic 3.0 T brain MRI findings in patients with aspartylglucosaminuria (AGU), a rare lysosomal storage disorder. Previous AGU patient material imaged at 1.0 and 1.5 T was also re-evaluated. Materials and methods: Twenty-five brain MRI examinations from 20 AGU patients were included in the study. Thirteen patients underwent a prospective 3.0 T MRI (5 male, 8 female, aged 9—45 years). Twelve examinations from nine patients (4 male, 5 female, aged 8—33 years) previously imaged at 1.0 or 1.5 T were re-evaluated. Two patients were included in both the prospective and the retrospective groups. Visual analysis of the T1- and T2-weighted images was performed by two radiologists. Results: The previously reported signal intensity changes in T2-weighted images were visible at all field strengths, but they were more distinct at 3.0 T than at 1.0 or 1.5 T. These included signal intensity decrease in the thalami and especially in the pulvinar nuclei, periventricular signal intensity increase and juxtacortical high signal foci. Poor differentiation between gray and white matter was found in all patients. Some degree of cerebral and/or cerebellar atrophy and mild ventricular dilatation were found in nearly all patients. This study also disclosed various unspecific findings, including a higher than normal incidence of dilated perivascular spaces, arachnoid cysts, pineal cysts and mildly dilated cavum veli interpositi. Conclusion: This study revealed particular brain MRI findings in AGU, which can raise the suspicion of this rare disease in clinical practice. © 2015 Elsevier Masson SAS. All rights reserved.

Abbreviations: MRI, magnetic resonance imaging; AGU, aspartylglucosaminuria; T, Tesla; AGA, aspartylglucosaminidase; SI, signal intensity; WM, white matter; GM, grey matter; T2 TSE, T2-weighted turbo spin echo; T1 3D TFE, T1-weighted three-dimensional turbo field echo; TR, repetition time; TE, echo time; SE, spin echo; T1 3D MP-RAGE, T1-weighted three-dimensional magnetization-prepared rapid acquisition gradient echo; FLAIR, fluid-attenuated inversion recovery; PVS, perivascular space; MCM, mega cisterna magna; CJD, Creutzfeldt—Jakob disease; CSF, cerebrospinal fluid; CVI, cavum veli interpositi. ∗ Corresponding author. Tel.: +358 9 4711. E-mail addresses: anna.tokola@hus.fi (A.M. Tokola), laura.aberg@fimnet.fi (L.E. Åberg), taina.autti@hus.fi (T.H. Autti). http://dx.doi.org/10.1016/j.neurad.2015.03.003 0150-9861/© 2015 Elsevier Masson SAS. All rights reserved.

Please cite this article in press as: Tokola AM, et al. Brain MRI findings in aspartylglucosaminuria. J Neuroradiol (2015), http://dx.doi.org/10.1016/j.neurad.2015.03.003

+Model NEURAD-515; No. of Pages 13

ARTICLE IN PRESS

2

A.M. Tokola et al.

Introduction Aspartylglucosaminuria (AGU) is an inherited, autosomal recessive, progressive neurodegenerative disease. It is a lysosomal storage disorder and manifests with progressive decline in cognitive and motor functions, leading to premature death [1]. The disease is caused by a mutation of the AGA gene, located on 4q34.3 [2], resulting in deficient activity of aspartylglucosaminidase, a lysosomal hydrolase enzyme. As a result, there is excessive accumulation of uncleaved glycoasparagines in lysosomes and elevated metabolite levels in urine. In Finland, the estimated incidence of the disease is 1:18,000, while in other countries, the incidence is unknown [3,4]. Sporadic cases are found in many countries and at least 27 different mutations of the gene exist [5]. Children with AGU appear healthy at birth. Delayed speech and clumsiness are typically the first symptoms noticed around the age of 2 to 3 years. Later on, progressive intellectual impairment becomes evident [6]. Typical facial features (a broad lower jaw, a short, broad nose, periorbital puffiness and round cheeks) become more pronounced during the teens [7]. Epilepsy may develop in adulthood and life expectancy is typically less than 45 years [1,6]. Lysosomal storage disorders that primarily affect the central nervous system have no known treatment available to cure or slow down the progression of the disease. Mouse model studies investigating virus-mediated gene therapy and enzyme replacement therapy in AGU have shown some metabolic correction and decreased storage in the brain [8—10]. Previous brain MRI studies at 1.0 and 1.5 T have shown thalamic T2 SI decrease, delayed myelination and increased T2 SI in the periventricular WM in AGU [1,11,12]. Furthermore, slow progressive brain atrophy has been reported. No specific SI alterations in T1-weighted images have been reported in AGU. A histopathologic study has revealed that the cortical and deep GM neurons contain vacuoles, and there is diffuse pallor of myelin staining and gliosis in the WM [1]. The aim of this study was to identify characteristic 3.0 T brain MRI findings in patients with AGU. Also, previous AGU patient material imaged at 1.0 and 1.5 T was re-evaluated to assess the impact of field strength on the findings.

Materials and methods In total, 25 brain MRI examinations from 20 AGU patients were included in the study. Thirteen patients (5 male, 8 female, age between 9 and 45 years, mean 25.5 years, SD 9.8) were recruited for this study, and underwent prospective 3.0 T MRI. In addition, retrospective MRI material from nine AGU patients (4 male, 5 female, age between 8 and 33 years, mean 15.6, SD 7.3) consisting of 12 examinations imaged at 1.0 T and/or 1.5 T was re-evaluated. Mean age was calculated using age at the first examination. One patient was imaged four times (twice at 1.0, once at 1.5 and 3.0 T) and two patients twice (one at 1.0 and 1.5 T, and the other at 1.5 and 3.0 T). In all cases, the AGU diagnosis was confirmed with a urine test showing elevated levels of aspartylglucosamin and a blood test showing deficiency in the AGA

enzyme. The patients imaged at 3.0 T are presented in Table 1, and the patients imaged at 1.0 and 1.5 T in Table 2. The patients imaged at 3.0 T were recruited with assistance from Suomen AGU Ry patient association and Rinnekoti Foundation. Informed consent was obtained from the parents of the patients prior to inclusion in the study. A clinical neurological examination was performed on all patients and data on their medical history were collected from parents and medical records. No sedation was used during the scanning and all patients were in good physical condition at the time of the examination. All patients were intellectually disabled and four of the patients had epilepsy. The study was approved by the ethics committee of the Hospital for Children and Adolescents, University of Helsinki.

MRI acquisition For the new patient material, the MR imaging was performed using a 3.0 T Achieva device (Philips, Best, Netherlands). The examination included a T2 TSE axial series (TR 4000, TE 80, slice thickness 4 mm, flip angle 90◦ , matrix 512 × 512) and a T1 3D TFE series (TR 8.19, TE 3.79, slice thickness 1 mm, flip angle 8◦ , matrix 256 × 256). From the previous patient material, five examinations were performed using a 1.0 T Magnetom Harmony device (Siemens, Erlangen, Germany) including T2 dual SE axial and coronal images (TR 2500, TE 90, slice thickness 5 mm, gap 1 mm, matrix 256 × 256) and T1 axial images (TR 600, TE 15, slice thickness 4 mm, gap 0.8 mm, matrix 256 × 256). Four examinations were performed using a 1.5 T Magnetom Vision device (Siemens, Erlangen, Germany) including a T2 TSE axial series (TR 3500 or 3000, TE 90 or 85, slice thickness 5 mm, gap 1 mm, flip angle 180◦ , matrix 256 × 256), a T1 3D MP-RAGE series (TR 1900, TE 3.68, TI 1100, flip angle 15◦ , matrix 512 × 512) and a FLAIR axial series (TR 9000, TE 105, TI 2500, flip angle 180◦ , slice thickness 5 mm, gap 1 mm, matrix 256 × 256). The remaining three examinations were performed using a 1.5 T Sonata device (Siemens, Erlangen, Germany) including a T2 TSE axial series (TR 6000, TE 125, slice thickness 5 mm, gap 1 mm, flip angle 150◦ and matrix 512 × 384) and a T1 3D MP-RAGE series (TR 1900, TE 3.68, TI 1100, flip angle 15◦ , matrix 512 × 512). The patients imaged at 1.0 and 1.5 T had participated in previous studies [11,12]. Their images were subsequently re-evaluated using the same parameters as in the 3.0 T group.

Visual analysis The T1 and T2-weighted images were evaluated by two radiologists. A FLAIR series was also available from four patients imaged at 1.5 T. The consensus of opinion was used as the final assessment. From the MR examinations, twelve parameters were evaluated. These included thalamic (1) and pulvinar nuclei SI changes (2), periventricular (3) and other white matter SI abnormalities (4), GM/WM differentiation (5), PVS dilatation (6), widening of cerebral sulci (7) and cerebellar fissures (8), enlargement of ventricles (9), abnormalities of the corpus callosum (10), cystic changes (11) and the presence of cavum variations (12).

Please cite this article in press as: Tokola AM, et al. Brain MRI findings in aspartylglucosaminuria. J Neuroradiol (2015), http://dx.doi.org/10.1016/j.neurad.2015.03.003

Sex

Thalamus T2 SI

Pulvinar T2 SI

Periventricular Juxtacortical GM/WM T2 SI T2 foci differentiation

Cerebral and cerebellar atrophy

CC atrophy

PVS dilatation

Other

1

12.0

F

Decreased

Decreased

Increased

None

Poor

Some prominent sulci

Atrophy

Normal

2

17.3

F

Decreased

Decreased

Normal

None

Poor

Atrophy

Normal

3

26.6

F

Decreased

Decreased

Mildly increased

Several

Poor

Mild cerebral and cerebellar atrophy Mild cerebral and cerebellar atrophy

Pineal cyst 8 mm, mild CVI dilatation 13 mm, lateral ventricles borderline normal Lateral ventricle dilatation

Atrophy

Dilatation ≤ 2 mm

4

30.4

F

Decreased

Decreased

Increased

Several

Poor

Mild cerebral and evident cerebellar atrophy

Atrophy

Dilatation > 2 mm

5

32.4

F

Decreased

Decreased

Increased

Several

Poor

Mild cerebral atrophy

Atrophy

Dilatation > 2 mm

6

33.5

F

Heterogenous/ Heterogenous/ Increased increased increased

Several

Poor

Mild cerebral and evident cerebellar atrophy

Atrophy

Dilatation ≤ 2 mm

Pineal cyst 6 mm, CVI dilatation 12 mm, lateral and 3rd ventricle dilatation Pineal cyst 6 mm, choroid plexus cysts 10 mm and 13 mm, CSP, mild lateral and 3rd ventricle dilatation Pineal cyst 5 mm, small retrocerebellar arachnoid cyst or MCM, CVI dilatation 11 mm, mild lateral ventricle dilatation Multilocular pineal cyst 14 mm, mild lateral and 3rd ventricle dilatation

ARTICLE IN PRESS

Age

+Model

Patient

NEURAD-515; No. of Pages 13

Patients imaged with 3.0 T and their findings.

Brain MRI findings in aspartylglucosaminuria

Please cite this article in press as: Tokola AM, et al. Brain MRI findings in aspartylglucosaminuria. J Neuroradiol (2015), http://dx.doi.org/10.1016/j.neurad.2015.03.003

Table 1

3

Thalamus T2 SI

Pulvinar T2 SI

Periventricular Juxtacortical GM/WM T2 SI T2 foci differentiation

Cerebral and cerebellar atrophy

CC atrophy

PVS dilatation

Other

7

34.7

F

Decreased

Decreased

Mildly increased

Several

Poor

Mild cedebral and evident cerebellar atrophy

Atrophy

Dilatation ≤ 2 mm

8

45.3

F

Heterogenous/ Decreased decreased

Increased

Several

Poor

Evident cerebral and cerebellar atrophy

Atrophy

Dilatation ≤ 2 mm

9

9.9

M

Decreased

Decreased

Increased

None

Poor

Mild cerebral and cerebellar atrophy

Atrophy

Dilatation ≤ 2 mm

10

19.5

M

Decreased

Heterogenous/ Mildly decreased increased

<5

Poor

Mild cerebral and cerebellar atrophy

Borderline normal

Normal

11

21.8

M

Decreased

Decreased

<5

Poor

Mild cerebral and cerebellar atrophy

Borderline normal

Normal

Pineal cyst 9 mm, choroid plexus cysts 5 mm, mild and 3rd ventricle dilatation Prominent propably multilocular corpus pineale, CVI dilatation 10 mm, lateral and 3rd ventricle dilatation, some disturbing motion artifact Pineal multilocular cyst 6 mm, CVI dilatation 16 mm, mild lateral ventricle dilatation Retrocerebellar arachnoid cyst or MCM 25 mm, choroid plexus cysts 5 mm, some motion artifact, mild lateral and 3rd ventricle dilatation Choroid plexus cyst 6 mm, mild lateral and 3rd ventricle dilatation

Increased

ARTICLE IN PRESS

Sex

+Model

Age

A.M. Tokola et al.

Patient

NEURAD-515; No. of Pages 13

4

Please cite this article in press as: Tokola AM, et al. Brain MRI findings in aspartylglucosaminuria. J Neuroradiol (2015), http://dx.doi.org/10.1016/j.neurad.2015.03.003

Table 1 (Continued)

Thalamus T2 SI

Pulvinar T2 SI

Periventricular Juxtacortical GM/WM T2 SI T2 foci differentiation

Cerebral and cerebellar atrophy

CC atrophy

PVS dilatation

Other

12

22.5

M

Decreased

Decreased

Mildly increased

<5

Poor

Evident cerebral and cerebellar atrophy

Atrophy

Dilatation > 2 mm

13

25.9

M

Decreased

Decreased

Increased

<5

Poor

Evident cerebral and mild cerebellar atrophy

Atrophy

Dilatation ≤ 2 mm

Pineal cyst 5 mm, choroid plexus cysts 6 mm and 6 mm, temporal arachnoid cysts 49 mm and 14 mm, retrocerebellar arachnoid cyst 39 mm or MCM, CVI dilatation 10 mm, lateral and 3rd ventricle dilatation Pineal cyst 5 mm, choroid plexus cysts 8 mm and 9 mm, retrocerebellar arachnoid cyst 56 mm or MCM, lateral and 3rd ventricle dilatation

F: female; M: male; SI: signal intensity; GM: grey matter; WM: white matter; CC: corpus callosum; PVS: Virchow-Robin perivascular spaces; CVI: cavum veli interpositi; CSP: cavum septi pellucidi; MCM: mega cisterna magna. Patient number 2 is also imaged with 1.0 and 1.5 T (patient number 1 in that group) and patient number 3 is also imaged with 1.5 T (patient number 7 in that group).

ARTICLE IN PRESS

Sex

+Model

Age

NEURAD-515; No. of Pages 13

Patient

Brain MRI findings in aspartylglucosaminuria

Please cite this article in press as: Tokola AM, et al. Brain MRI findings in aspartylglucosaminuria. J Neuroradiol (2015), http://dx.doi.org/10.1016/j.neurad.2015.03.003

Table 1 (Continued)

5

Patient

Age

Thalamus T2 SI

Pulvinar T2 SI

Periventricular Juxtacortical T2 SI T2 foci

GM/WM differentiation

Cerebral and cerebellar atrophy

CC atrophy

PVS dilatation

Other

8.7

F

1.0

Decreased

Decreased

Mildly increased

None

Poor

Atrophy

Normal

10.8

F

1.5

Decreased

Decreased

Normal

None

Poor

Atrophy

Normal

13.7

F

1.5

Decreased

Decreased

Normal

None

Poor

Atrophy

Normal

9.2

M

1.0

Decreased

Decreased

Increased

None

Poor

Mild cerebral atrophy Mild cerebral and vermis atrophy Mild cerebral and cerebellar atrophy Mild cerebral and vermis atrophy

Atrophy

Normal

16.3

M

1.5

Decreased

Decreased

Increased

<5

Poor

Mild cerebral and cerebellar atrophy

Atrophy

Dilatation ≤ 2 mm

3

9.9

F

1.0

Decreased

Decreased

Increased

None

Poor

Borderline normal

Normal

4

10.2

F

1.5

Decreased

Decreased

Increased

<5

Poor

Atrophy

Dilatation > 2 mm

5

14.7

M

1.5

Decreased

Decreased

Increased

Several

Poor

Some prominent cerebral sulci Mild cerebral and vermis atrophy Mild cerebral and cerebellar atrophy

Lateral and 3rd vetricle dilatation Lateral and 3rd ventricle dilatation, no change Lateral and 3rd ventricle dilatation, no change Retrocerebellar arachnoid cyst or MCM, mild lateral ventricle dilatation Retrocerebellar arachnoid cyst or MCM, mild lateral ventricle dilatation, no change Mild CVI dilatation

Atrophy

Dilatation > 2 mm

1

2

Choroid plexus cysts 5 mm and 7 mm, mild CVI dilatation

ARTICLE IN PRESS

Tesla

A.M. Tokola et al.

Sex

+Model

Patients imaged with 1.0 and 1.5 T and their findings.

NEURAD-515; No. of Pages 13

6

Please cite this article in press as: Tokola AM, et al. Brain MRI findings in aspartylglucosaminuria. J Neuroradiol (2015), http://dx.doi.org/10.1016/j.neurad.2015.03.003

Table 2

Tesla

Thalamus T2 SI

Pulvinar T2 SI

Periventricular Juxtacortical T2 SI T2 foci

GM/WM differentiation

Cerebral and cerebellar atrophy

CC atrophy

PVS dilatation

Other

6

16.1

F

1.0

Decreased

Decreased

Increased

None

Poor

Atrophy

Dilatation < 2 mm

7

16.5

F

1.5

Decreased

Decreased

Increased

Several

Poor

Mild cerebral and vermis atrophy Mild cerebral and cerebellar atrophy

Atrophy

Normal

8

21.0

M

1.5

Decreased

Decreased

Increased

Several

Poor

Mild cerebral atrophy

Atrophy

Normal

9

33.0

M

1.0

Decreased

Decreased

Increased

Several

Poor

Evident cerebral and mild cerebellar atrophy

Atrophy

Normal

Mild lateral and 3rd ventricle dilatation, mild CVI dilatation Pineal multilocular cyst, choroid plexus cysts 5 mm and 5 mm, mild CVI dilatation, mild lateral ventricle dilatation Retrocerebellar arachnoid cyst or MCM and temporal arachnoid cysts, mild CVI dilatation, lateral and 3rd ventricle dilatation Mild CVI dilatation, lateral ventricle dilatation

F: female; M: male; SI: signal intensity; GM: grey matter; WM: white matter; CC: corpus callosum; PVS: Virchow-Robin perivascular spaces; MCM: mega cisterna magna; CVI: cavum veli interpositi; CSV: cavum septum vergae. Patients number 1 and 7 are also imaged with 3.0 T (patients number 2 and 3, respectively).

ARTICLE IN PRESS

Sex

+Model

Age

NEURAD-515; No. of Pages 13

Patient

Brain MRI findings in aspartylglucosaminuria

Please cite this article in press as: Tokola AM, et al. Brain MRI findings in aspartylglucosaminuria. J Neuroradiol (2015), http://dx.doi.org/10.1016/j.neurad.2015.03.003

Table 2 (Continued)

7

+Model NEURAD-515; No. of Pages 13

ARTICLE IN PRESS

8

A.M. Tokola et al.

Results The SI changes in T2-weighted images were distinct at 3.0 T. As expected, the image quality and tissue contrast were better at 3.0 T than at 1.0 and 1.5 T. The decreased SI in the thalami was evident, and especially at 3.0 T the decrease was more intense in the dorsal area of the thalami, i.e., the pulvinar nuclei (Fig. 1). Decreased SI in the thalami was found in 12 out of 13 (92%) patients imaged at 3.0 T and it was more intense in the pulvinar nuclei in 11 (85%) patients. One patient had motion artifacts in the images and the thalami were evaluated as abnormally heterogeneous, but with decreased SI also in the pulvinar nuclei. One patient had an atypical heterogeneous T2 SI increase in the thalami, and especially in the pulvinar nuclei bilaterally (Fig. 2). In T1-weighted images the pulvinar nuclei showed decreased SI. In addition, this patient had more patchy T2 SI increase foci in the periventricular and juxtacortical WM than the other patients. All patients imaged at 1.0 or 1.5 T showed decreased T2 SI in the thalami, but the focal SI decrease in the pulvinar nuclei did not visualize as evidently as at 3.0 T (Figs. 1 and 3). Altogether, 19 out of 20 patients (95%) showed decreased SI in the thalami and the pulvinar nuclei in the T2-weighted images. An increased SI in the periventricular WM in T2-weighted images was also found in all patients, except for a female patient who was imaged four times. Increased periventricular SI in this patient’s T2-weighted images was found only in the first images taken at the age of 8 years at 1.0 T. Patchy juxtacortical foci with increased SI in T2-weighted images were found in 10 out of 13 (77%) patients imaged with 3.0 T; 6 of these patients had more than five foci (Fig. 3). Juxtacortical foci were not noted in the images of the three youngest patients (age 9.9 and 12.0 and 17.3). Altogether, 15 out of 20 (75%) patients spanning both groups had juxtacortical high SI foci, and in both groups the oldest patients had several foci. Decreased contrast between the cerebral cortex and underlying WM was found in all patients with AGU. Cerebral atrophy was found in 18 out of 20 (90%) patients, which was evaluated as mild in most cases. The images of one young patient from both groups were evaluated as borderline normal with some prominent cortical sulci (age 9.9 and 12.0 years). The three oldest patients (aged 22.5, 25.9 and 45.3 years) in the 3.0 T group and the oldest patient imaged with 1.0 T (age 33.0 years) had more pronounced cerebral atrophy. A total of 16 out of 20 (80%) patients had some cerebellar atrophy, but 11 of these had only mild cerebellar or vermis atrophy. Five patients (aged 22.5, 30.4, 33.5, 34.7 and 45.3 years) in the 3.0 T group had more pronounced cerebellar atrophy. The three patients that were imaged more than once showed progression in both cerebral and cerebellar atrophy over time. Mild lateral ventricle dilatation was a common feature; it was found in 18 out of 20 (90%) patients and mild third ventricle dilatation was noticed in 12 out of 20 (60%) patients (Figs. 1—4). Interestingly, thin corpus callosum was evident in 17 out of 20 (85%) patients (Fig. 4). In the remaining three patients, the corpus callosum was evaluated as borderline normal. Also, it was noted that all patients had a relatively thick skull.

Figure 1 Female aspartylglucosaminuria patient, imaged three times, axial T2-weighted images. Top row, at 8 years of age imaged at 1.0 T; second row, at 13 years of age imaged at 1.5 T; and third row, at 17 years of age imaged at 3.0 T. The typical T2 hypointensity in the thalami and especially in the pulvinar nuclei is more distinctly visualized at 3.0 T than at 1.0 and 1.5 T. Also, mild general atrophy, mild lateral and 3rd ventricle dilatation and poor differentiation between grey and white matter is noted. The signal intensity in the globi pallidi is also low compared with a control participant. Last row, healthy control participant, a 17-year-old female, imaged at 3.0 T showing normal signal intensity in the thalami and normal myelination.

Please cite this article in press as: Tokola AM, et al. Brain MRI findings in aspartylglucosaminuria. J Neuroradiol (2015), http://dx.doi.org/10.1016/j.neurad.2015.03.003

+Model NEURAD-515; No. of Pages 13

ARTICLE IN PRESS

Brain MRI findings in aspartylglucosaminuria

9

Figure 2 A 33-year-old female aspartylglucosaminuria patient, imaged at 3.0 T. On the left are four axial T2-weighted images showing atypical signal intensity increase in the thalami that has not previously been reported in aspartylglucosaminuria, and several patchy foci with signal intensity increase in the periventricular and juxtacortical white matter. On the right are two sagittal T1weighted images showing respectively SI decrease in the pulvinar nucleus and some focal hypointensity lesions in the white matter. Also, mild lateral and 3rd ventricle dilatation, poor differentiation between grey and white matter, mildly dilated perivascular spaces, a pineal cyst and mild general atrophy are visible in the images of this patient.

Figure 3 Two aspartylglucosaminuria patients imaged at 3.0 T. On the left is a 34-year-old female patient and on the right is a 26-year-old female patient. Axial T2-weighted images showing typical signal intensity decrease in the pulvinar nuclei. In both, poor differentiation between grey and white matter, some patchy juxtacortical foci with signal intensity increase, mild dilatation of the 3rd ventricle, a pineal cyst and a relatively thick skull are also visible.

Four examinations imaged with 1.5 T included a FLAIR sequence. In these, the SI in the thalami was decreased especially in the pulvinar nuclei, and mild periventricular SI increase was also noted (Fig. 5). Prominent PVSs were found in nine out of 13 (69%) patients imaged at 3.0 T (Figs. 2 and 4). Altogether, enlarged

PVSs were visualized in 13 out of 20 (65%) patients. The image quality, especially in the 1.0 T images, may not be sufficient to demonstrate mildly dilated PVSs. Several cystic changes were also discovered (Figs. 1—4). Pineal cysts measuring from 5 to 9 mm were found in nine out of 13 (69%) patients imaged at 3.0 T and in 10 out of 20 patients (50%) in total. Choroid plexus cysts were also a common finding. These cysts, measuring from 5 to 13 mm, were found in six out of 13 (46%) patients imaged at 3.0 T and in eight out of 20 (40%) patients in total. Also, six out of 20 patients (30%) had retrocerebellar cystic changes, either arachnoid cysts or mega cisterna magna. In addition, temporal arachnoid cysts were noticed in two patients, one of whom also had a retrocerebellar arachnoid cyst or MCM. Mildly dilated, 10 mm or more wide cavum veli interpositi were found in 11 out of 20 patients (55%), (Fig. 5). One patient had cavum septum vergae variation. The patients imaged at 3.0 T are presented in Table 1, and the patients imaged at 1.0 and 1.5 T in Table 2. Table 3 displays a summary of brain MRI findings in all 20 patients with AGU in our study.

Discussion For this study, we have gathered 25 brain MRI examinations from 20 AGU patients, which is to our knowledge the largest MRI study material published so far in this rare disease. This

Please cite this article in press as: Tokola AM, et al. Brain MRI findings in aspartylglucosaminuria. J Neuroradiol (2015), http://dx.doi.org/10.1016/j.neurad.2015.03.003

+Model NEURAD-515; No. of Pages 13

ARTICLE IN PRESS

10

A.M. Tokola et al. Table 3 Summary of brain MRI findings in 20 aspartylglucosaminuria patients. Brain MRI finding

Frequency, %

Decreased T2 SI in the thalami Decreased T2 SI in the pulvinar nuclei Increased T2 SI in periventricular WM Poor differentiation between GM and WM Juxtacortical foci with increased T2 SI Cerebral atrophy, mostly mild Cerebellar atrophy, mostly mild Corpus callosum hypoplasia/atrophy Mild lateral ventricle dilatation Mild 3rd ventricle dilatation Prominent PVSs Pineal cysts Arachnoid cysts or MCM

95 95 100 100 75 90 80 85 90 60 65 50 35

SI: signal intensity; WM: white matter; GM: grey matter; PVS: Virchow-Robin perivascular spaces; MCM: mega cisterna magna.

Figure 4 Two aspartylglucosaminuria patients imaged at 3.0 T. The top row shows a 22-year-old male patient, both sagittal T1-weighted images, and the second row shows a 25year-old male patient, with one axial T2-weighted image and one sagittal T1-weighted image. In both patients, a thin corpus callosum, mild general atrophy and a retrocerebellar cyst (arachnoid cyst or mega cisterna magna) is noted. Also, on the images in the top row, a temporal arachnoid cyst and mildly dilated perivascular spaces are visible.

is also the first study evaluating brain MRI findings in AGU at 3.0 T. 3.0 T field strength provides a higher signal-to-noise ratio than 1.0 and 1.5 T. Furthermore, 3.0 T imaging is potentially more sensitive in lesion detection in the brain. The tissue contrast between GM and WM is better, which makes evaluation of myelination and gliosis more distinct. Likewise, a better resolution reveals small structural changes, i.e., enlarged PVSs and small cystic changes more clearly. The retrospective material was imaged several years ago and the 1.0 T images in particular were of rather poor quality. Modern equipment with 1.5 T field strength has better

image quality and the typical findings in AGU are well evaluable at 1.5 T. The magnetic field strength may have substantial influence on SI alterations caused by storage material. Although the precise mechanism causing the SI alterations in AGU is not known, it has been suggested that the accumulation of macromolecules and lipids can increase viscosity in tissue, shorten both T1 and T2 times and therefore, cause the typical SI decrease in T2-weighted images in the thalami [11]. Bilateral thalamic lesions are seen in various disorders, including metabolic, neoplastic, vascular and inflammatory disorders and infections [13]. In some neurodegenerative diseases, such as multiple sclerosis, Alzheimer’s disease and Huntington’s disease, the T2 hypointensity of the basal ganglia and thalamus is considered to be related to iron accumulation [14,15]. Previous studies have also shown thalamic SI changes in many lysosomal storage disorders, but among these AGU is the only one with marked T2 hypointensity in the pulvinar nuclei without signal alterations on other weightings [11,12,16]. In visual analysis, the stronger magnetic field seems to accentuate the T2 hypointensity in the thalami and especially the pulvinar nuclei.

Figure 5 A 10-year-old female aspartylglucosaminuria patient, imaged at 1.5 T. Fluid-attenuated inversion recovery (FLAIR) axial images showing signal intensity decrease in the pulvinar nuclei and mild signal intensity increase in the periventricular white matter.

Please cite this article in press as: Tokola AM, et al. Brain MRI findings in aspartylglucosaminuria. J Neuroradiol (2015), http://dx.doi.org/10.1016/j.neurad.2015.03.003

+Model NEURAD-515; No. of Pages 13

ARTICLE IN PRESS

Brain MRI findings in aspartylglucosaminuria The mean age of the patients in our study imaged at 1.0 and 1.5 T was lower than that of those imaged at 3.0 T. However, the decreased T2 SI in the thalami and the pulvinar nuclei was remarkable even in the youngest patients imaged at 3.0 T. As for the thalamic findings, there was one outlier in the 3.0 T patient group. A female patient showed bilateral T2 SI increase in the thalami that was emphasized in the pulvinar nuclei (Fig. 2). In T1-weighted images the SI was decreased. This rare finding has not previously been associated with AGU. High T2 SI in the pulvinar nuclei is described as a typical sign in variant Creutzfeldt—Jakob disease and it can also occur in sporadic CJD [13,17]. Our patient was only diagnosed with AGU and the clinical picture was typical. In AGU, the bilateral T2 SI increase in the thalami and the pulvinar nuclei may be associated with demyelination or gliosis, but the pathophysiologic mechanism is unclear. A FLAIR sequence was only available from four patients imaged with 1.5 T. It has been previously reported, that the SI decrease in the pulvinar nuclei is clearly visible in FLAIR sequence [12]. In addition, mild periventricular SI increase in the WM is visible on FLAIR images (Fig. 5). It was also noted that in some cases in AGU, the globi pallidi have to some extent lower T2 SI than in healthy individuals (Fig. 1). However, it is difficult to evaluate this reliably with visual analysis, as the T2 signal in the globi pallidi is also low in healthy individuals. In a previous histopathology study it was found that vacuolation in the neurons of the basal ganglia and thalami was most severe in the globi pallidi and the lateral parts of the thalami [1]. More work needs to be done to evaluate the significance of this finding. In AGU, excessive accumulation of uncleaved glycoasparagines in lysosomes probably leads to dysmyelination or demyelination, gliosis and neuronal loss. Previous imaging studies have concentrated on GM changes, and the thalamic SI decrease in T2-weighted images is emphasized. With 3.0 T the WM changes, including periventricular T2 SI increase and patchy juxtacortical foci, are more appreciable. Poor differentiation between GM and WM was found in all patients, indicating hypomyelination, demyelination and/or gliosis. The juxtacortical high SI foci in T2-weighted images that were found in all but the two youngest patients imaged at 3.0 T are also considered a sign of demyelination or gliosis affecting the U-fibers. The periventricular SI increase in T2weighted images also represents dysmyelination. Enlarged PVSs are a nonspecific finding and they are often also found in healthy participants [18]. At high resolution 3.0 T MR, 25—30% of healthy children demonstrate some prominent PVSs [18], but these are usually not dilated, i.e., not more than 2 mm wide. The prevalence of dilated PVSs in young adult population under 30 years of age has been reported to be 1.6% [19]. PVSs are pia-lined spaces that accompany penetrating arteries and arterioles into the brain parenchyma and are filled with interstitial fluid. Enlarged PVSs are found in many neurodegenerative disorders, including some lysosomal storage diseases [16,20,21]. Patients with Hurler, Hunter or Sanfilippo disease have been found to accumulate undegraded mucopolysaccarides within enlarged PVSs, but it has also been hypothesized that PVS dilatation may reflect the impairment of CSF reabsorption because of deposits in the leptomeninges [16]. Also, the

11 visibility and dilatation of PVSs has been associated with inflammatory and vascular diseases. In our study, at 3.0 T 69% of patients showed some prominent PVSs in the basal ganglia and the WM of the cerebrum, and in three of these patients (aged 22.5, 30.4 and 32.4 years) dilated PVSs measuring more than 2 mm were found. Enlarged PVSs have not previously been reported in AGU. The pathophysiology of slowly progressive cerebral and cerebellar atrophy in AGU remains unclear. Accumulation, neuronal death and gliosis are probable causes, but impaired reabsorption of CSF resulting in ventriculomegaly and the enlargement of subarachnoid spaces might contribute to this. Furthermore, corpus callosum thinning appears to be a typical finding in AGU and it was also noted in the youngest patients in this study. The corpus callosum is a major WM bundle connecting the hemispheres, primarily the homologous areas of the cortex. Its atrophy is found in many neurodegenerative diseases, such as Alzheimer’s disease, frontotemporal dementia and multiple sclerosis, and it has also been associated with cognitive decline and executive deficits in patients with age-related WM hyperintensities [22—25]. Moreover, corpus callosum hypoplasia or atrophy has been described as a typical feature in some lysosomal storage diseases [16,26,27]. In AGU, corpus callosum hypoplasia or atrophy may be explained with hypomyelination, demyelination and gliosis, and delayed transmission of information between different brain areas via corpus callosum tracts can have a significant role in the clinical picture. Arachnoid cysts are congenital intra-arachnoidal lesions filled with CSF. They may be secondary to splitting of the developing arachnoid or in the middle fossa as a consequence of the failure of temporal embryonic meninges to merge as the sylvian fissure forms. Also, fluid secretion by the cyst wall, trauma, meningitis and subarachnoid hemorrhage have all been suggested as causes [18,28]. Arachnoid cysts are relatively common, and are found in about 2% of patients as an incidental finding [28]. In our study, 30% of patients had retrocerebellar cystic changes of which some may have been MCM, and 10% had temporal arachnoid cysts. Therefore, it appears that the incidence of arachnoid cysts in AGU patients exceeds the incidence in the normal population. Posterior fossa arachnoid cysts and MCM have also been reported in other lysosomal storage disorders, and disturbance of CSF circulation is among the explanatory theories [16]. Pineal cysts and cystic degeneration of the pineal gland are also a common finding in imaging studies. In routine imaging they are seen in up to 23% of individuals and even more at autopsy [28]. Theories regarding the origin of pineal cysts include ischemic glial degeneration with or without hemorrhagic expansion, enlargement of pre-existing cysts under hormonal influence and end enlargement of the embryonic pineal cavity. In our 3.0 T study group nine out of 13 patients (69%) had pineal cysts, which is more than average in this age group. Furthermore, choroid plexus cysts were frequent in our AGU patient material, as these were noted in 40% of individuals. However, choroid plexus cysts are such a common finding in the healthy population that this cannot be considered abnormal or remarkable. According to the literature, choroid plexus cysts occur in up to

Please cite this article in press as: Tokola AM, et al. Brain MRI findings in aspartylglucosaminuria. J Neuroradiol (2015), http://dx.doi.org/10.1016/j.neurad.2015.03.003

+Model NEURAD-515; No. of Pages 13

ARTICLE IN PRESS

12 50% of autopsies, although the proportion is not as high in adolescents [28]. CVI is a normal variation where the cistern of the cavum veli interpositi is dilated. This CSF-containing space lies between the velum interpositum and the inferior aspect of the corpus callosum, and it communicates directly with the quadrigeminal cistern [29]. CVI is a frequent imaging finding in infants and young children (up to 34%), but in most cases it closes in early childhood and the incidence in children over 2 years of age is 2—3%. In the adult population the incidence is reported to be between 5.5 to 7.2% [30]. In our study, 55% of AGU patients had mildly dilated CVI measuring more than 10 mm in width. CVI is usually considered an incidental finding, but an association with intellectual disability, epilepsy and autism has also been reported and large CVI may cause obstructive hydrocephalus [30,31]. None of the AGU patients in our study showed large CVI and no symptoms of obstructive hydrocephalus were noted. The CVI dilatation in AGU is probably related to atrophy and ventricular dilatation. Cavum septum vergae variation was found in only one patient. The limitation of this study was the small number of patients, which is related to the rarity of the disease. The imaging protocol and imager were the same in all 3.0 T examinations, which minimized variations based on technical reasons.

Conclusion Although AGU is a rare disease, sporadic cases exist worldwide. Therefore, when there is a question of unexplained intellectual disability, typical brain MRI findings can raise a suspicion of AGU. The most typical finding is decreased T2 SI in the thalami and especially in the pulvinar nuclei. A combination of more unspecific findings including high T2 SI in the periventricular WM, poor differentiation between GM and WM, thin corpus callosum, some degree of cerebral and/or cerebellar atrophy and mild ventricular dilatation was also found in nearly all patients with AGU in this study. The cerebral and cerebellar atrophy seems to progress with age. Furthermore, patchy juxtacortical high SI foci in T2-weighted images are a common finding in this disorder. This study also uncovered several unspecific findings in AGU patients that have previously been described in other lysosomal storage disorders and various neurodegenerative diseases. Therefore, a higher than normal incidence of prominent PVSs, arachnoid cysts, pineal cysts and mildly dilated CVI also seem to be associated with AGU.

Ethical standards and patient consent We declare that all human studies have been approved by the ethics committee of the Hospital for Children and Adolescents, University of Helsinki, and have therefore been performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments. We declare that all patients and parents of the patients gave informed consent prior to inclusion in this study.

A.M. Tokola et al.

Disclosure of interest The authors declare that they have no conflicts of interest concerning this article.

Acknowledgments The authors wish to thank the families attending the study and the Finnish AGU Association/Suomen AGU Ry. This study was supported by Rinnekoti Research Fund/Finnish Brain Foundation, Arvo and Lea Ylppö Foundation, Yrjö Jahnsson Foundation, Pehr Oscar Klingendahl Fund and the Scientific Fund of HUS Medical Imaging Center of Helsinki University Central Hospital. Part of the results has been presented at the ESNR meeting 2012.

References [1] Autti T, Raininko R, Haltia M, et al. Aspartylglucosaminuria: radiologic course of the disease with histopathologic correlation. J Child Neurol 1997;12:369—75. [2] Saarela J, Laine M, Oinonen C, et al. Molecular pathogenesis of a disease: structural consequences of aspartylglucosaminuria mutations. Hum Mol Genet 2001;10:983—95. [3] Ikonen E, Peltonen L. Mutations causing aspartylglucosaminuria (AGU): a lysosomal accumulation disease. Hum Mutat 1992;1:361—5. [4] Hietala M, Gron K, Syvanen AC, et al. Prospects of carrier screening of aspartylglucosaminuria in Finland. Eur J Hum Genet 1993;1:296—300. [5] Saarela J, von Schantz C, Peltonen L, et al. A novel aspartylglucosaminuria mutation affects translocation of aspartylglucosaminidase. Hum Mutat 2004;24:350—1. [6] Arvio P, Arvio M. Progressive nature of aspartylglucosaminuria. Acta Paediatr 2002;91:255—7. [7] Arvio MA, Peippo MM, Arvio PJ, et al. Dysmorphic facial features in aspartylglucosaminuria patients and carriers. Clin Dysmorphol 2004;13:11—5. [8] Dunder U, Valtonen P, Kelo E, et al. Early initiation of enzyme replacement therapy improves metabolic correction in the brain tissue of aspartylglycosaminuria mice. J Inherit Metab Dis 2010;33:611—7, http://dx.doi.org/10.1007/ s10545-010-9158-7. [9] Virta S, Rapola J, Jalanko A, et al. Use of nonviral promoters in adenovirus-mediated gene therapy: reduction of lysosomal storage in the aspartylglucosaminuria mouse. J Gene Med 2006;8:699—706. [10] Kelo E [Dissertation] Catalytic and therapeutic charasteristics of human recombinant glycosylasparaginase and bacterial L-asparaginases. Faculty of Health Sciences, University of Eastern Finland; 2013. [11] Autti T, Joensuu R, Aberg L. Decreased T2 signal in the thalami may be a sign of lysosomal storage disease. Neuroradiology 2007;49:571—8. [12] Autti T, Lonnqvist T, Joensuu R. Bilateral pulvinar signal intensity decrease on T2-weighted images in patients with aspartylglucosaminuria. Acta Radiol 2008;49:687—92. [13] Smith AB, Smirniotopoulos JG, Rushing EJ, et al. Bilateral thalamic lesions. AJR Am J Roentgenol 2009;192:W53—62, http://dx.doi.org/10.2214/AJR.08.1585. [14] Zecca L, Youdim MB, Riederer P, et al. Iron, brain ageing and neurodegenerative disorders. Nat Rev Neurosci 2004;5:863—73.

Please cite this article in press as: Tokola AM, et al. Brain MRI findings in aspartylglucosaminuria. J Neuroradiol (2015), http://dx.doi.org/10.1016/j.neurad.2015.03.003

+Model NEURAD-515; No. of Pages 13

ARTICLE IN PRESS

Brain MRI findings in aspartylglucosaminuria [15] Walsh AJ, Blevins G, Lebel RM, et al. Longitudinal MR imaging of iron in multiple sclerosis: an imaging marker of disease. Radiology 2014;270:186—96, http://dx.doi.org/10.1148/ radiol.13130474. [16] Zafeiriou DI, Batzios SP. Brain and spinal MR imaging findings in mucopolysaccharidoses: a review. AJNR Am J Neuroradiol 2013;34:5—13, http://dx.doi.org/10.3174/ajnr.A2832. [17] Zeidler M, Sellar RJ, Collie DA, et al. The pulvinar sign on magnetic resonance imaging in variant Creutzfeldt-Jakob disease. Lancet 2000;355:1412—8. [18] Osborn AG, Preece MT. Intracranial cysts: radiologicpathologic correlation and imaging approach. Radiology 2006;239:650—64. [19] Groeschel S, Chong WK, Surtees R, et al. Virchow-Robin spaces on magnetic resonance images: normative data, their dilatation, and a review of the literature. Neuroradiology 2006;48:745—54. [20] Calleja Gero ML, Gonzalez Gutierrez-Solana L, Lopez Marin L, et al. Neuroimaging findings in patient series with mucopolysaccharidosis. Neurologia 2012;27:407—13. [21] Patlas M, Shapira MY, Nagler A, et al. MRI of mannosidosis. Neuroradiology 2001;43:941—3. [22] Ryberg C, Rostrup E, Paulson OB, et al. Corpus callosum atrophy as a predictor of age-related cognitive and motor impairment: a 3-year follow-up of the LADIS study cohort. J Neurol Sci 2011;307:100—5, http://dx.doi.org/10.1016/ j.jns.2011.05.002.

13 [23] Ryberg C, Rostrup E, Sjostrand K, et al. White matter changes contribute to corpus callosum atrophy in the elderly: the LADIS study. AJNR Am J Neuroradiol 2008;29:1498—504, http://dx.doi.org/10.3174/ajnr.A1169. [24] Meguro K, Constans JM, Shimada M, et al. Corpus callosum atrophy, white matter lesions, and frontal executive dysfunction in normal aging and Alzheimer’s disease. A community-based study: the Tajiri project. Int Psychogeriatr 2003;15:9—25. [25] Dietemann JL, Beigelman C, Rumbach L, et al. Multiple sclerosis and corpus callosum atrophy: relationship of MRI findings to clinical data. Neuroradiology 1988;30:478—80. [26] Sonninen P, Autti T, Varho T, et al. Brain involvement in Salla disease. AJNR Am J Neuroradiol 1999;20:433—43. [27] Frei KP, Patronas NJ, Crutchfield KE, et al. Mucolipidosis type IV: characteristic MRI findings. Neurology 1998;51:565—9. [28] Osborn AG. Nonneoplastic cysts. In: Osborn’s brain; imaging, pathology, and anatomy. 1st ed. Canada: Amirsys, Inc.; 2013. p. 773—807. [29] Osborn AG. Hydrocephalus and CSF disorders, normal variants. In: Osborn’s brain; imaging, pathology, and anatomy. 1st ed. Canada: Amirsys, Inc.; 2013. p. 1012. [30] Saba L, Anzidei M, Raz E, et al. MR and CT of brain’s cava. J Neuroimaging 2013;23:326—35, http://dx.doi.org/10.1111/ jon.12004. [31] Kier. The evolutionary and embryologic basis for the development and anatomy of the cavum veli interpositi. AJNR Am J Neuroradiol 1999;20:1383—4.

Please cite this article in press as: Tokola AM, et al. Brain MRI findings in aspartylglucosaminuria. J Neuroradiol (2015), http://dx.doi.org/10.1016/j.neurad.2015.03.003