Enricomaria Mormina1&Marilena Briguglio2&Rosa Morabito1&Alessandro Arrigo1&
Silvia Marino3&Gabriella Di Rosa2&Alessia Micalizzi4,5&Enza Maria Valente4,6&
Vincenzo Salpietro2&Sergio Lucio Vinci1&Marcello Longo1&Francesca Granata1
# Springer Science+Business Media New York 2015
Abstract Aim of this study is to show the potential of prob- abilistic tractographic techniques, based on the Constrained Spherical Deconvolution (CSD) algorithms, in recognizing white matter fiber bundle anomalies in patients with complex cerebral malformations, such as cerebellar agenesis. The mor- phological and tractographic study of a 17-year-old male pa- tient affected by cerebellar agenesis was performed by using a 3Tesla MRI scanner. Genetic and neuropsychological tests were carried out. An MRI morphological study showed the absence of both cerebellar hemispheres and the flattening of the anterior side of the pons. Moreover, it showed a severe vermian hypoplasia with a minimal vermian residual. The study recognized two thin cerebellar remnants, medially in contact with the small vermian residual, at the pontine level. The third ventricle, morphologically normal, communicated with a permagna cerebello-medullary cistern. Probabilistic CSD tractography identified some abnormal and aberrant infratentorial tracts, symmetrical on both sides. In particular, the transverse pontine fibers were absent and the following tracts with aberrant trajectories have been identified: Bcerebello-thalamic^ tracts; Bfronto-cerebellar^ tracts; and ip- silateral and contralateralBspino-cerebellar^ tracts. Abnormal tracts connecting the two thin cerebellar remnants have also been detected. There were no visible alterations in the main
supratentorial tracts in either side. Neuropsychiatric evalua- tion showed moderate cognitive-motor impairment with dis- crete adaptive compensation. Probabilistic CSD tractography is a promising technique that overcome reconstruction biases of other diffusion tensor-based approaches and allowed us to recognize, in a patient with cerebellar agenesis, abnormal tracts and aberrant trajectories of normally existing tracts.
Keywords Probabilistic . Tractography . Cerebellar agenesis . Cerebellar hypoplasia . Hindbrain malformation . Pons malformation
Background
Congenital structural abnormalities of the cerebellum are rare conditions commonly associated with neurodevelopmental disorders. Posterior fossa malformations, such as cerebellar hypoplasia, pontine tegmental cap dysplasia, or Joubert syn- drome, can be caused by several genetic mutations and are known to be related to neurocognitive and behavioral dys- functions in varying degrees (Steinlin2008; Briguglio et al.
2011; Brancati et al.2010). Furthermore cerebellar abnormal- ities could also be acquired when a disruption consequence
* Alessandro Arrigo
alessandro.arrigo@hotmail.com
1
Neuroradiology Unit - Department of Biomedical Sciences and Morpho-Functional Imaging, University of Messina, via Consolare Valeria, 1 A.O.U. PoliclinicoBG. Martino^, 98125 Messina, Italy
2 Department of Pediatric, Gynecological, Microbiological and
Biomedical Sciences, University of Messina, Messina, Italy
3 IRCCS Centro Neurolesi Bonino-Pulejo, Messina, Italy 4
IRCCS Casa Sollievo della Sofferenza, CSS-Mendel Laboratory, San Giovanni Rotondo, Italy
5
Department of Biological and Environmental Science, University of Messina, Messina, Italy
6 Section of Neurosciences, Department of Medicine and Surgery,
University of Salerno, Salerno, Italy Brain Imaging and Behavior
occurrs in the perinatal period, commonly due to extreme prematurity (less than 32 weeks of gestational age), or to a cytomegalovirus infection (Messerschmidt et al. 2005; Poretti et al.2009; Namavar et al.2011a,b; Poretti et al.
2014).
Previous studies paid great attention to the role of cerebel- lar structural abnormalities in the etiopathogenesis of neuro- psychological dysfunctions and a relationship between early development of cognitive functions and extent of vermian lobulation has been reported (Boddaert et al.2003; Basson and Wingate2013). Some authors also focused on synaptic reorganization, which might establish functional compensa- tion after cerebellar structural damage (Sotelo and Privat
1978). These high plasticity properties of the cerebellum were especially detected in younger brains, after structural cerebel- lar alterations (Chugani et al. 1996) and, as suggested by Bolduc and Limperopoulos (2009), in these pathological con- ditions, anBearly targeted intervention could potentially trans- late into reorganization of the cerebellar circuitry and result in improved outcome^.
Advanced neuroimaging studies should be encouraged in patients with cerebellar abnormalities in order to investigate possible anatomical or functional pathways, allowing partial compensation of the damage of the closed loop cerebro- cerebellar circuit, which is currently considered the main cer- ebellar output (Strick et al.2009; Granata et al.2013). Fiber tractography, an advanced neuroimaging technique based on Magnetic Resonance Imaging (MRI), can be an essential tool in both evaluation and detection of white matter fiber bundle alterations in several cerebral structural abnormalities (Poretti et al.2007; Jissendi-Tchofo et al.2009; Wahl et al. 2010; Poretti et al. 2013; Jissendi-Tchofo et al. 2014; Huisman et al.2014). The aim of our work is to show the tractographic findings of a rare case of cerebellar agenesis and the utility of this advanced MRI technique in understanding the abnormal neuroanatomy of this particular condition.
Patient and methods
The entire study was approved by our Ethical Committee. The patient’s history and neurodevelopmental data were obtained through parents’ interviews.
Neuropsychological assessment
Neuropsychological battery included: non-verbal Scale Leiter-R, Raven’s progressive matrices, Vineland Behaviour Scale (VABS), Child Behavior Checklist scale (CBCL 6–18), Visuo-Motor Integration test (VMI) and full scale intelligence quotient (IQ).
Genetic testing
Karyotype was determined using standard procedures. The GeneChip 6.0 platform (Affymetrix, Santa Clara, CA) was used to analyze patients’ genomic DNA following manufac- turer’s instructions. Copy number analysis was performed with the Genotyping Console v4.0 (Affymetrix).
Mutation analysis of a panel of 44 genes causative of non- progressive forms of congenital ataxias (with the exclusion of Joubert syndrome genes) was performed with the True-Seq Custom Amplicon (TSCA) technology (Illumina), and an Illumina MiSeq platform, following the well-established Illumina protocol. The panel included the PTF1A gene, as well as known genes responsible for various forms of ponto- cerebellar hypoplasia (TSEN54, TSEN34, TSEN2, VRK1, EXOSC3, RARS2, AMPD2, CASK, CHMP1A, SEPSECS) and of whole cerebellar hypoplasia (OPHN1).
MRI acquisition protocol
All MR images were performed by means of a 3T MR scanner (Achieva 3.0T Philips, Best, The Netherlands) by using an 8- channel SENSE phased-array radio-frequency head coil. Morphological MR imaging of the brain has been obtained by multiplanar T1-weighted MPR-GE sequence, T2-weighted high-resolution sagittal and coronal FSE images, FLAIR, Double Inversion Recovery (Grey Matter only). Diffusion MR imaging has been obtained with a dual phase-encoded pulsed-gradient spin-echo diffusion weighted imaging se- quence, performed with 60 non-collinear directions and a b- value of 1000 s/mm2.
Pre-processing and tractographic reconstructions Motion and eddy current distortion correction of the diffusion weighted images were performed with toolboxes available within SPM8 (www.fil.ion.ucl.ak.uk/spm).
Probabilistic tractography was performed on the pre- processed diffusion weighted images, with a probabilistic Constrained Spherical Deconvolution (CSD) algorithm (Tournier et al. 2007; Mormina et al.2015) by using the MRtrix package (JD Tournier, Brain Research Institute, Melbourne, Australia,http://www.brain.org.au/software).
Results
Clinical, neurological, and genetic features
The patient is a 17-year-old male, born from non- consanguineous healthy parents. No family history of cerebral nervous system malformations was reported. All clinical, neu- rological, and genetic findings are summarized in Table1.
Table 1 Clinical data, neuropsychological assessments, and genetic findings of our patient, from birth to present study Age Category Data description
Before / at birth Pregnancy medical history Negative tests for TORCH agents
No maternal use of any drugs or teratogens agents exposure No alcohol or cocaine use by the mother
Delivery medical history At 35 weeks of gestational age Premature rupture of membranes Birth weight: 2,08 Kg (<3rd percentile) Hypotonia, weak cry, diffuse cyanosis
<3 years Developmental milestones Smile at 7 months; hold control at 8 months; sit and babbling at 12 months
3 years Motor skills Independent walk achieved (ataxic and uncoordinated); gross and fine motor impairment; ocular motor dyspraxia, horizontal nystagmus and strabismus; dysfagia and oromotor dyspraxia; sphincters not controlled
Speech skills Expressive language consisted of a few single words Relatively spared comprehension
Social behavior Social and relational skills moderately impaired Dysmorphic features Head
- Fronto-orbital constriction
- Enlarged, flattened and asymmetric nose root - Short columella
Body
- Camptodactyly of the V finger of both hands
- Bilateral partial cutaneous syndactyly of the II, III, and IV toes - Mild degree of dorsal scoliosis
Skin
- Right submammary hypochromic stain - Cutaneous hyperlaxity
Follow-up every 6 months from 3 to 17 years
Motor skills Prominent dysmetria (well shown by finger to nose test)
Prominent dysarthria (with scanning speech and monotonous prosody) Fine motor skills moderately compromised
Speech skills Verbal skills mildly improved
Expansion of either receptive and expressive vocabulary Supportive therapy Developmental, speech, and special educational trainings 17 years (age of present study) Cognitive skills tests Non-verbal Leiter-R Scale
- Moderate intellectual dysability (IQ score of 44) - Best performance: fluid reasoning (score of 63) - Worst performance: visuo-spatial tasks (score of 59) Raven’s progressive matrices - Low scores (<3rd percentile) emerged
Neurolinguistic assessments Higher scores for receptive vocabulary and syntactic comprehension Mild deficit of expressive language, writing, and dictation
Visuoperceptual tasks Visual Motor Integration (VMI) test: score of 57 (1st percentile) Adaptive functioning tests Vineland Adaptive Behavior Scale (VABS)
- Really good daily living skills in home environment and near complete daily self care activities, such as: personal hygiene, dressing, undressing, eating, and voluntary control of sphincters
Behavior tests Child Behavior CheckList (CBCL) 6–18 test - T est 64; T int 59; T tot 68
- Low scores in attention and complex thinking subscales - Prevalent externalizing behavior problems
Genetic tests No karyotype alterations were found
No mutations were found on these genes: PTF1A, TSEN54, TSEN34, TSEN2, VRK1, EXOSC3, RARS2, AMPD2, CASK, CHMP1A, SEPSECS, OPHN1
Other tests Echocardiography and abdominal ultrasound scan did not show any anomaly Blood tests excluded endocrine or metabolic disorders Brain Imaging and Behavior
Morphological MRI
Morphological MRI examination mainly detected the complete absence of both cerebellar hemispheres and severe vermian hypoplasia (Fig.1a). No annular protuberance, which was more evident in sagittal slices (Fig.1b), has been found. The third ventricle was regularly shaped, communicating by a nor- mal Sylvian aqueduct with a sort of permagna cerebello- medullary cistern filled by cerebrospinal fluid. The cistern was shaped like the missing cerebellum, with no profiles or volume alterations of the posterior fossa nor tentorium position- ing abnormalities (Fig. 1b). Two structures with dead-end, which appeared as cerebellar remnants, have been found at the pontine level (Fig.1c, d and e), connected to each other by a thin lamina. A small median structure, resembling a thin vermian residual, was also connected to this lamina and located dorso-caudally to the quadrigeminal lamina (Fig. 1b). Hyperintense signal was found in cerebellar remnants (Fig.1e).
There was no hydrocephalus and all the supratentorial structures, including diencephalon and telencephalon, showed entirely normal morphology.
CSD fiber tractography
The main supratentorial fiber bundles were bilaterally evalu- ated and no morphological abnormalities have been found out
(Fig.2). We detected different fiber bundle anomalies in the infratentorial region, such as abnormal tracts and normally existing tracts with aberrant trajectories (Figs.3,4,5and6). All tractographic findings are summarized in Table2.
Discussion
The morphological MRI examination of our patient showed a rare complex structural alteration of cranial posterior fossa, characterized by cerebellar agenesis (intended as absence of both cerebellar hemispheres), severe vermian hypoplasia, and the lack of the annular protuberance of the pons. Previous studies described similar conditions of cerebellar agenesis in several patients, with a huge spectrum of morphological and clinical findings (Yoshida and Nakamura 1982; Sener and Jinkins 1993; Sener 1995; Van Hoof and Wilmink 1996; Velioglu et al.1998; Deniz et al.2002; Timmann et al.2003; Poretti et al.2008, 2009; Namavar et al.2011a,b; Yu et al.
2014). For this reason, different classifications have been pro- posed to describe this pathological condition and its variants, such as profound cerebellar hypoplasia or primary cerebellar agenesis– with total and partial absence of the cerebellum – and type IV Chiari malformation (Barkovich et al. 2009; Sener and Jinkins 1993; Yu et al.2014). In our paper, we preferred the term Bcerebellar agenesis^, as suggested by
Fig. 1 Morphological Magnetic Resonance images of the rare case studied. a Axial MPR GE T1- weighted image shows the absence of cerebellum in the posterior fossa (asterisk) and the vermian residual (arrow). b Sagittal MPR GE T1-weighted image shows the vermian residual (arrow), the flattening of the anterior side of the pons (empty arrows), the absence of cerebellum (asterisk) and the normal positioning of the tentorium (arrowheads). c Axial MPR GE T1-weighted and d FSE T2-weighted images show the cerebellar remnants (arrows) and the lack of cerebellar hemispheres (asterisks). e Axial Double Inversion Recovery shows heterotopic grey matter signal in the cerebellar remnants (arrows)
Poretti et al. (2009), to mean a pathological condition in which only minute cerebellar tissue is detectable, without taking into account the underlying pathomechanism.
Based on conventional MRI, cerebellar agenesis or near total absence of the cerebellum is presumably linked to PTF1A gene mutation, whereas unilateral or asymmetrical cerebellar hypoplasia/agenesis is likely due to a disruptive prenatal event, extreme prematurity, or acquired causes lead- ing to cerebellar disruption (Messerschmidt et al.2005,2008a,
b; Poretti et al.2008,2009; Namavar et al.2011a,b). The term Bcerebellar disruption^ was used by Messerschmidt et al. (2005)Bto emphasize the combination of acquired damage after preterm birth^, due to prenatal infections in the cerebellar vulnerability period, with a subsequent developmental failure. It is also known that poor postnatal conditions and cerebellar hemosiderin deposits– as a result of hemorrhagic lesion – are perinatal high-risk factors for cerebellar disruption (Messerschmidt et al.2008a, b). On these bases, cerebellar agenesis is considered the most severe form of cerebellar dis- ruption, occurring during pregnancy as a bilateral disruptive event, caused by infections, ischemia, or hemorrhage (Poretti et al.2009; Jissendi-Tchofo et al.2014). Nevertheless, al- though the disruptive hypothesis partially matches with our case (cerebellar agenesis with no genes mutation found and
delivery problems), we could not confirm or exclude it as the main pathomechanism, due to the lack of extreme prematurity and diseases occurring during pregnancy (see Table1for de- tails). On the other hand, as a possible alternative pathomechanism, we might take into account the hypothesis of a not yet classified gene mutation that could lead to a pri- mary malformation.
Further, as yet reported by Yu and colleagues (Yu et al.
2014), there are only poorly detailed descriptions of clinical and neurological features of cerebellar agenesis. Based on the growing role of the cerebellum on cognitive processes and emotional control, in addition to its role in motor coordination (Schmahmann and Sherman 1998; Schmahmann 2004; Schmahmann and Caplan2006), patients with neurocognitive and behavioral dysfunctions and cerebellar malformations should be more commonly reported.
The neuropsychological evaluation of our proband re- vealed less severe cognitive and motor impairments with good compensation in motor, cognitive, affective, and relational skills. These could appear disproportionate considering the expected outcome for the morphological neuroimaging find- ings, but similar discrepancies are also reported in other cases of cerebellar agenesis (Sener and Jinkins1993; Sener1995; Yu et al.2014).
Fig. 2 Tractographic reconstruction of corticospinal tracts and arcuate fasciculi. a Coronal and b sagittal images show corticospinal tract reconstruction without any morphological abnormalities. c Sagittal and d tilted axial images depict normal arcuate fasciculi Brain Imaging and Behavior
Based on these data, we could hypothesize that during the growth of the subject, some complex cerebral circuits inter- vened as a compensation mechanism for cerebellar agenesis.
We performed MRI tractography in order to reveal the an- atomical substrate of our patient cerebral connectivity and eventually to recognize any significant white matter tract
modification that might help to understand the cerebellar agenesis compensation.
Several studies reported how tractography could be helpful in the specific tracking of white matter tracts in some pathological conditions, such as: Joubert syndrome, pontine tegmental cap dysplasia, and the agenesis of the corpus callosum (Jissendi-
Fig. 4 Tractographic reconstruction of tracts connecting the cerebellar remnants. a Coronal and b axial views of tractographic reconstructions of the fiber bundles connecting the two thin cerebellar remnants. c Axial morphological view with the overlay of the color-coded FA map, showing the vermian residual (arrow) colored in red due to the latero-lateral passage of tracts connecting the cerebellar remnants
Fig. 3 Tractographic reconstruction ofBspino- cerebellar^ tracts. a Right and b left ipsilateral spino-cerebellar tracts are showed in coronal view. c and d images show respectively axial view of both ipsilateral spino-cerebellar tractographic reconstructions and the morphological axial image at the same level. e Left and f right contralateral spino-cerebellar tracts are showed in coronal view. g The decussation of these tracts is showed in detail in axial view. Sagittal rotated images show respectively h the cerebellar and vermian remnants (arrow), and i the tractographic reconstruction of the decussation of contralateral spino-cerebellar tracts (arrow)
Tchofo et al.2009; Wahl et al.2009; Wahl et al.2010; Jissendi- Tchofo et al.2014). Previous tractographic reports of patients with ponto-cerebellar hypoplasia and cerebellar agenesis, per- formed by means of deterministic diffusion tensor imaging (DTI) tractography, showed missing cerebellar afferent or effer- ent fibers (Yu et al.2014) and the absence of transverse pontine fibers (Poretti et al.2013). Although our results confirmed the absence of transverse pontine fibers, we were also able to find several Bcerebellar^ tract projections to other brain areas, infratentorial tract abnormalities and aberrant trajectories of
normally existing tracts, which were not found in previous stud- ies performed with commonly used DTI tractographic tech- niques (Poretti et al.2013; Yu et al.2014).
In our work, for the white matter tract evaluation of our patient, we chose a probabilistic CSD-based tractographic technique, due to its inherent ability to overcome the well- known pitfalls of the commonly used deterministic DTI algo- rithms. This technique is able to better evaluate voxels con- tainingBcrossing^ fibers (Tournier et al.2008), and allowed us to identify tracts with aberrant trajectories and abnormal tracts
Fig. 5 Tractographic reconstruction ofBcerebello- thalamic^ tracts. a Coronal, b axial and c sagittal rotated views show theBcerebello-thalamic^ tracts (arrows), connecting the cerebellar residuals with thalami (T)
Fig. 6 Tractographic reconstruction ofBfronto- cerebellar^ tracts. a Axial tilted and b sagittal views show bilateralBfronto-cerebellar^ tracts. c Left and d right fronto- cerebellar tracts are showed in axial tilted views, with both ipsilateral and contralateral components
(summarized in Table2). The tractographic study of our pa- tient revealed some normally existing tracts with aberrant tra- jectories, such as: ipsilateral and contralateral Bspino- cerebellar^ tracts, Bcerebello-thalamic^ and Bfronto- cerebellar^ tracts. The course evaluation of these tracts (Table2), compared with the usual course of correspondent tracts in healthy subjects (Brodal2010; van Baarsen et al.
2013; Soelva et al.2013), allowed us to reveal abnormal tracts and some significant additional information on the complex anatomical condition of cerebellar peduncles, which were not readily recognizable with morphological imaging. In particu- lar, we might recognize the middle cerebellar peduncle resid- uals in the small cerebellar remnants, which are connected each other by some abnormal tracts (Fig.4) and the inferior cerebellar peduncle residuals in the thin structures connecting the spinal cord with cerebellar remnants. Moreover, the eval- uation of color coded FA maps showed the absence of pontine transverse fibers and the presence of superior cerebellar pe- duncles with their decussation (seen as the usualBred dot^, which indicate a latero-lateral fibers passage). The latter find- ing was confirmed by fiber tractography, which revealed that some tracts crossed the midline at the level of superior cere- bellar peduncle decussation (Table2).
We could speculate that these tracts might represent a part of the anatomical substrate resulting in functional
compensation for the lack of cerebellum and pontine hypopla- sia. Hence, we would recommend the use of probabilistic CSD-based tractography for the examination of this type of disease, in order to better detect white matter tract course anomalies.
Nevertheless, further studies combining probabilistic tractographic approaches and functional MRI could be helpful to understand the mechanisms leading to sufficient motor, cognitive, and affective compensations, which might occur