Review. The role of magnetic resonance imaging in elucidating the pathogenesis of cerebral palsy: a systematic review

Transcription

1 The role of magnetic resonance imaging in elucidating the pathogenesis of cerebral palsy: a systematic review Review Ingeborg Krägeloh-Mann* MD, Professor of Paediatrics; Veronka Horber MD, University Children s Hospital, Department of Paediatric Neurology, Tuebingen, Germany. *Correspondence to first author at University Children s Hospital, Department of Paediatric Neurology, Hoppe- Seyler-Str1, D Tuebingen, Germany. ingeborg.kraegeloh-mann@med.uni-tuebingen.de The aim of this study was to show the role of magnetic resonance imaging (MRI) in elucidating the aetiology, or at least pathogenesis, of cerebral palsy (CP). A systematic review of studies using MRI in children with CP was performed according to pathogenetic patterns characterizing different timing periods of occurence of the lesions, and with respect to gestational age (term vs preterm) and CP subtypes. Out of the studies published since 1990 in English, six met all the inclusion criteria; they involved children with spastic and dyskinetic CP. Abnormal MRI was reported in 334 out of 388 (86%) patients and gave clues to pathogenesis in 83%. Fourteen studies met only part of the inclusion criteria and abnormal MRIs were reported even more frequently in these (91%; 930/1022). Periventricular white matter lesions were most frequent (56%) followed by cortical and deep grey matter lesions (18%); brain maldevelopments were rather rare, described in 9%. Brain maldevelopments and grey matter lesions were more often seen in term than in pretermborn children with CP (brain maldevelopments: 16% vs 2.5%; grey matter lesions: 33% vs 3.5%); periventricular white matter lesions occurred significantly more often in preterm than in term-born children (90% vs 20%). CP is mainly characterized by brain lesions which can be identified by MRI in around 75% of preterm infants; brain maldevelopments occur in around 10%. Cerebral palsy (CP) is one of the most common forms of severe childhood disability and it is especially related to preterm birth. CP prevalence is around 2 per 1000 live births and increases to 40 to 100 per 1000 live births among infants born very early or with very-low birthweight. 1,2 CP brings with it many associated problems with learning, language, and epilepsy, as well as hearing and vision impairment. 3 The definition is usually based on phenomenology; it only specifies that CP originates from an interference, lesion, or abnormality of the developing brain. Origin before the end of the neonatal period is usually distinguished from a postneonatal origin. 4 Neuroimaging, especially magnetic resonance imaging (MRI), plays an increasing role in the diagnosis of CP. 5 It has the potential to visualize physiological and pathological morphological changes during brain development. 6 The human brain undergoes complex organizational changes during development, in and ex utero. Pathogenic events affecting the developing brain cause abnormalities/lesions, the patterns of which depend on the stage of brain development. Cortical neurogenesis takes place predominantly during the 1st and 2nd trimester and is characterized by proliferation, migration, and organization of neuronal precursor cells, then neuronal cells. Brain pathology is characterized by maldevelopments caused by genetic or acquired impairments. 7,8 From the late 2nd and early 3rd trimester onwards, when the grossarchitecture of the brain (neural cyto- and histogenesis) is established, growth and differentiation events (axonal and dendrite growth, synapse formation, and myelination) are predominant which persist into postnatal life. Disturbances of brain development during this period most often result in lesions/defects. Their causes are multiple and key factors are inflammation with excessive cytokine production, oxidative stress, and excess release of glutamate, triggering the excitotoxic cascade, factors 144 Developmental Medicine & Child Neurology 2007, 49:

2 which are induced by infectious and hypoxic-ischemic mechanisms A potentiation of single effects can be assumed. 12,13 During the early 3rd trimester, and in the preterm-born infant, periventricular white matter (PWM) is especially affected. Towards the end of the 3rd trimester, and in the term-born infant, grey matter, either cortical or deep grey matter (e.g. basal ganglia and thalamus) appear to be more vulnerable Infarcts of the middle cerebral artery (MCA) are reported mainly in term or near-term born infants, 20,21 although they may occur in the very preterm infant. 22 Thus, different patterns of brain abnormalities (whether maldevelopments or lesions in the sense of defects) characterize different timing periods of compromise. Taking into account that CP originates from an interference, lesion, or abnormality of the developing brain, it can be assumed that MRI has the potential to identify the lesion or abnormality and, thus, can help us to understand the timing of CP origin. The aim of this review is to show the role of MRI in elucidating the aetiology, or at least the pathogenesis, of CP by analyzing MRI results according to the aetio-pathogenetic patterns indicated above. An emphasis was placed on analyzing term- and preterm-born infants separately as the higher prevalence in preterm children may indicate a different pathogenesis. Methods A literature search was done for relevant papers in English language from 1990 to October 2006, using the key terms MRI, cerebral palsy, spastic diplegia, spastic tetraplegia, and spastic hemiplegia. Articles were excluded if the aim of the MRI was other than to establish the aetiology or pathogenesis of CP (e.g. papers analyzing only one of the above mentioned patterns, or papers dealing with brain morphometry only). Papers were included if they met the following criteria: (1) Inclusion of patients according to the clinical diagnosis of CP. Only cases with a clear neurology (spastic, dyskinetic also dystonic or choreoathetoid and ataxic CP) were considered. Hypotonic CP was excluded. (2) Age at least 2 years at clinical diagnosis of CP. A definite diagnosis of CP in the infant and Bilateral A B C D Conception 6wks 20wks 30wks 40wks birth a b c d Unilateral Brain maldevelopments Periventricular lesions Grey matter lesions 1st 2nd trimester or genetic early 3rd trimester late 3rd trimester 9% 56% 18% Figure 1: Cerebral palsy (CP) and pathogenetic patterns, numbers in % come from group 1a and indicate, how often these patterns were described in CP according to the systematic review (ataxic CP excluded). Typical examples are given for different timing periods in uni- or bilateral (US, BS) distribution (lower, upper row respectively). Upper row from left: A, lissencephaly-pachygyria, T 1 -weighted (T1w) axial, 15-year-old male, BS-CP; B, Schizencephaly, T1w axial, 12-year-old male, BS-CP; C, bilateral periventricular leukomalacia (PVL) with gliosis (arrows) and white matter reduction, 6-year-old child, BS-CP; D, basal-ganglia/thalamus lesions (combined arrows) and central cortical lesions (arrows), T2w axial, 8 -year-old child, BS-CP. Lower row from left pathology is always left on image: a, hemimegalencephaly, T1w axial, 12- month-old male, US-CP; b, cortical dysplasia, T1w axial, 16-year-old male, US-CP; c, nearly unilateral PVL with gliosis (arrow), T2w axial, 6-year-old, male, US-CP; d, infarct of medial artery, T2w axial, 22-year-old female, US-CP. Review 145

3 younger child may be difficult because of the transitory and changing nature of early neurological signs, 23,24 this is why CP registers only include children at 3 years of age and definitely at 5 years. 4 (SCPE 2000). We compromised on an age at last Per cent Normal Miscellaneous Grey matter lesions Periventricular lesions Maldevelopments Term 122 BS-CP Preterm 186 Term 44 US-CP Figure 2: MRI pattern distribution in two major CP subtypes. Bilateral spastic cerebral palsy (BS-CP) and unilateral spastic cerebral palsy (US-CP) according to gestational age grouping term/preterm. Preterm 14 clinical follow-up of at least 2 years, otherwise too many studies would have been excluded. (3) Age at least 1 year at MRI. The correct identification of a lesional pattern of the brain with gliosis may be difficult before the end of myelination. Although neonatal MRI is considered by some authors to be superior to ultrasound in detecting neonatal brain lesions, its relationship to neurological outcome is not yet clearly established. 11 We chose an age of at least 1 year at MRI, unless the pathology was obvious and clearly described. (4) MRI as neuroimaging method according to a standardized protocol or re-evaluated on the basis of the original imaging data. (5) Separation into term or preterm (gestational age <37 wks) groups. MRI results were classified according to the above aetiopathogenetic patterns (for additional and detailed description see Krägeloh-Mann 25 ). Three groups were differentiated using the following patterns: (1) brain maldevelopments or 1st and 2nd trimester patterns which are supposed to occur in utero, such as lissencephaly, pachygyria, cortical dysplasias, polymicrogyria, and schizencephaly; (2) PWM lesions which are related to the early 3rd trimester of pregnancy and the preterm born infant, such as periventricular leukomalacia (PVL) defects following intraventricular haemorrhage (IVH) or periventricular haemorrhagic infarctions; and (3) cortical or deep grey matter lesions which occur towards the end of gestation late 3rd trimester patterns and peri- or neonatally, such as basal ganglia/thalamus lesions, parasagittal injury, multicystic encephalomalacia; MCA infarcts are also included here. Patterns which were considered abnormal, but not meeting the criteria above, were classified as miscellaneous. CP classification was given according to the subtypes indicated in the papers. For global analysis, diplegia and tetraplegia were summarized as bilateral-spastic CP (BS-CP); spastic hemiplegia was called unilateral-spastic CP (US-CP) following the European classification, and choreo-athetoid and dystonic CP were summarized as dyskinetic CP. 26 CP of postneonatal origin was excluded. Table I: Papers fulfilling inclusion criteria; Group 1a includes 388 magnetic resonance imaging (MRI) results from spastic and dyskinetic cerebral palsy (CP) described as abnormal in 334 (86%); Group 1b is a population-based study on ataxic CP (in part MRI) Paper Number of Abnormal Age at Age at last CP Subtypes (n) Brain Periventricular participants a MRI MRI, y follow-up, y maldevelopments lesions Preterm Term Preterm Term Group 1a Koeda et al Diplegia (18) Yokochi et al Diplegia (34) Yokochi et al Athetoid (22) c Krägeloh-Mann et al. 30b >5 BS-CP (56) leg-dominated other Okumura et al a 139 a 1 19 >2 Diplegia (80) Tetraplegia (41) US-CP (26) Kwong et al a 81 a 1 16 >2 US-CP (32) Diplegia (61) Tetraplegia (18) Group 1b Esscher et al. 33b (CT + MRI)? (5 21) Ataxic CP a Postneonatal CP excluded (Okumura et al. included five and Kwong et al. 11 post-neonatal cases); b population based; c lesions in periventricular white matter only, adjacent to trigone and posterior body of lateral ventricles; d for normal MRI, separation into term and preterm not given in this paper and assumed to be similar to other papers. BS-CP, bilateral spastic cerebral palsy; US-CP, unilateral spastic cerebral palsy; CT, computed tomography. 146 Developmental Medicine & Child Neurology 2007, 49:

4 STATISTICAL ANALYSIS We analyzed the data using χ 2 ; p<0.05 was considered statistically significant. Results Three hundred and twenty-five references were reviewed out of which 21 papers qualified to be considered in more detail. Only six met the strict inclusion criteria (Group 1a) As these studies did not contain any ataxic CP we accepted one exception (Group 1b), 33 a population-based study with very detailed clinical description and standardized, blinded imaging analysis. However, imaging was only in part MRI, age at imaging could not clearly be identified, and grouping into preterm/term was not given. In order to enlarge the basis for evaluation we accepted in a second step 14 papers that met part of our inclusion criteria (Group II) Almost all the studies analyzed neuroimaging in hospitalbased groups of children with CP. There were only two population-based trials out of the 21 papers reviewed. 30,33 Table I: continued Grey matter lesions Miscellaneous Normal Preterm Term Preterm Term Preterm Term d d 3 43 GROUP 1A Five studies in this group were hospital based and one population based. Two studies considered all CP types, and four examined CP subtypes only (three BS-CP and subtypes, one athetoid CP). There were a total of 388 patients (excluding 16 patients of postneonatal CP origin). Before pooling the data, we addressed the question of comparability. The population-based study on BS-CP 30 reported similar results to the other studies when considering BS-CP; abnormal results were reported in the first in 53 out of 56 cases (94%), which was slightly higher than the 224 out of 252 (89%) in the other studies, this was, however, not significantly different (p>0.05; Table I). MRI showed abnormalities in 334 (86%) cases; in 322 cases (83%) MRI results could be clearly allocated to the above mentioned patterns. In 3% findings were miscellaneous (Fig. 1). Two hundred (52%) of the 388 children were born preterm. Distribution into CP subgroups was: 308 (79%) BS-CP, 58 (15%) US-CP, and 22 (6%) athetoid. ANALYSIS ACCORDING TO CP SUBTYPES In children with BS-CP, 267 out of 308 (87%) had abnormal MRI findings. Brain maldevelopments were found in 27 (9%) and PWM lesions were observed in 193 (63%). Thirty-nine (13%) had cortical or deep grey matter lesions; and 8 (3%) could not be classified (miscellaneous). There was a clear difference between preterm- and term-born children with BS- CP (Fig. 2): PWM lesions occurred significantly more often in preterm than in term BS-CP (167/186 vs 26/122, [90% vs 21%]; p<0.001). Brain maldevelopments and cortical or deep grey matter lesions occurred significantly less often in preterm than term BS-CP (brain maldevelopments: 3/186 vs 24/122 [1.5%, vs 20%] and cortical/deep grey matter lesions: 7/186 vs 32/122; [4% vs 26%]; p<0.001). In US-CP, 52 out of 58 patients (90%) had abnormal MRimaging. Brain maldevelopments accounted for 9 (16%) of the cases; PWM lesions were revealed in 21 children (36%); 18 (31%) had cortical or deep grey matter lesions and 4 (7%) were miscellaneous. Again, as in BS-CP, there was a clear difference between preterm and term US-CP (Fig. 2). PWM lesions occurred significantly more often in preterm US-CP than in term (12/14 vs 9/44, e.g. 86% vs 20%; p<0.001) and cortical or deep grey matter lesions significantly less often (0/14 vs 18/44, e.g. 0 vs 41%; p<0.01) In contrast to BS-CP, brain maldevelopments occurred in preterm US-CP nearly as often as in term US-CP (2/14 vs 7/ 44, [14% vs 16%]; p>0.05). Dyskinetic CP, specified as athetoid CP, was dealt with in one study and 15 out of 22 children had abnormal MRI-findings. Twelve out of 22 children had cortical or deep grey matter lesions, three had PWM lesions ANALYSIS ACCORDING TO MRI PATTERNS Brain maldevelopments were reported in 9% (36/388) of cases. They occurred significantly more often in term than in preterm-born children with CP (31/188 vs 5/200, e.g. 16% vs 2.5%, p<0.001). CP-subtypes were especially severe forms of BS-CP (19/36, e.g. 53%) and US-CP (9/36). PWM lesions were reported in 56% (217/388) of cases. They were especially found in children born preterm (90%, 179/200; in comparison to 20%, 38/188, in term-born; p<0.001). PWM lesions were either PVL or consequences of IVH, or both. CP-subtypes reported were spastic CP, mostly milder bilateral forms, e.g. diplegia or leg-dominated BS-CP (151/179, 84% preterm-born children with PWM lesions had this form of CP) and less often, severe forms of BS-CP (16 tetraplegias [9%] were in the group of 179 preterm children, or US-CP (12 [7%]). In term-born children with CP, periventricular lesions were reported in 20% (38/188) of cases. CP subtypes were variable: 26 [68%] had some form of BS-CP which was described as diplegia or leg-dominated BS-CP in 18 (47%) and as quadriplegia in 8 (21%). Nine had US-CP (24%) and three athetoid CP (8%). In the bilateral forms, MRI patterns in detail were most often PVL, in US-CP, either focal periventricular gliosis rather than asymmetrical PVL, or unilateral porencephalic ventricular enlargement (suggesting IVH sequelae). Cortical or deep grey matter lesions could be found in 18% (69/388) of cases. They occurred more frequently in term- Review 147

5 born than in preterm children (62/188 [33%]), vs 7/200, [3.5%]; p<0.001). CP subtypes in the 62 term-born children were mainly severe forms of BS-CP (27 [44%] had tetraplegia or quadriplegia ) and athetoid CP (12/62, [19%]). Their MRI patterns were basal ganglia/thalamus or bilateral cortico-subcortical lesions. Eighteen of these 62 term-born children had US-CP and their MRI revealed mainly infarcts. When cortical or deep grey matter lesions were reported in former pretermborn children with CP, they tended to be of higher gestational age (in the paper where GA subgroups were given). 30 GROUP 1B This group consisted of one population-based study on ataxic CP. Abnormal imaging results were reported in 39% (27/90 cases), thus, significantly less often than in the other CP forms (334/388 [86%], p<0.001). In 17% each (12 cases each) cerebellar malformations (allocated to brain maldevelopments) and lesions respectively were identified; the latter were not further specified. In three cases (4%) findings were reported, which according to our criteria, could be considered miscellaneous. GROUP 2 This group included 14 studies, all hospital based. Six studies considered all CP types and seven considered CP subtypes only (four BS-CP and subtypes, three US-CP one ataxic CP called non-progressive ataxia ). There were a total of 1022 patients (excluding 10 patients of post-neonatal origin). MRI showed abnormalities in 930 cases (91%). MRI data were, however, difficult to analyze according to the criteria discussed above. This was because the MRI was, in part, performed early or could not be allocated to the patterns given above. CP definition often was not clear (atonic, hypotonic, unclassified, etc.) and age at clinical diagnosis was, in part, early or unclear. In one study, a multicentre European study, 47 analysis in term-versus preterm-born cases was not possible for the entire group, and CP subtype distribution could not be allocated to the MRI results. The total CP group had, however, similar results to the data given above, e.g. a predominance of lesional patterns with 71.7% of MRI results. PWM lesions were found in 42.5% of cases and cortical or deep grey matter lesions in 29.2%. Maldevelopments were seen in 9.1% of cases; normal findings were found in 11.7% and miscellaneous findings in 7.1% (Table II). Discussion A high percentage of abnormal MRIs were reported in all papers. In the reports of Group 1a, meeting all inclusion criteria and involving children with spastic and dyskinetic CP types, MRI was reported abnormal in 86% and gave clues to pathogenesis in 83% (excluding the 3% miscellaneous findings); normal or unspecific scans were rare. Only in Group 1b (ataxic CP) were abnormal findings less often seen (39%); ataxic CP is, however, rare and represents only around 4% of CP. 26 In Group II, with studies that met only parts of the inclusion criteria, abnormal MRIs were reported even more frequently (91%). Most of the studies were not population based and pretermborn children with CP were somewhat overrepresented (52%). In comparison, the Surveillance of Cerebral Palsy in Europe database, which is population based 4 gives 43% as the percentage of children with CP with a gestational age below 37 weeks, born between 1976 and 1996 (Cans C, personal communication, 2006). This probably also explains the fact that BS-CP was relatively more often present than US-CP (79% vs 15%, 6% being athetoid and ataxic CP not considered). In comparison, the subtype distribution in population-based studies is 55% for BS-CP, 30% for US-CP, 7% for dyskinetic, 4% for ataxic, with 4% being unknown. 26 The fact that studies meeting only part of the inclusion criteria reported abnormal MRI-findings to an even higher degree probably reflects a certain selection bias. Within studies of Group 1a (meeting all inclusion criteria) there was, however, no significant difference between the only population-based study and the hospital-based studies. Thus, we think that the analysis gives a rather reliable estimate of the clinical situation. PWM lesions were the most common pattern reported, in around 60% of children with CP (excluding ataxic CP) and by far the predominant MRI finding in preterm CP (approximately 90%). PVL and/or consequences of IVH of different topography and extent easily explain the CP-subtypes reported, e.g. spastic CP either bilateral and leg-dominated (diplegia) or unilateral. The fact that preterm CP has, in the vast majority, an injury pattern which corresponds to the period of birth of these children, does not tell us when exactly (shortly before, during, or after birth) the injury is acquired. This is scope for other studies, e.g. matched control studies involving pre-, peri-, and neonatal risk factors. CP-prevalence data could give indirect evidence as changes in prevalence in preterm CP reflects to 90% change in prevalence of this injury pattern. In the past 6 years there have been an increasing number of reports stating that CP prevalence in children with a birthweight of less than 1500g is declining, 2,48 51 although this is not seen in all countries. 52,53 PWM lesions were reported in around 20% of term-born children with CP, again mainly in milder BS-CP forms (diplegia), mostly a PVL pattern, but also in US-CP with focal periventricular gliosis rather than asymmetrical PVL or unilateral porencephalic ventricular enlargement (suggesting IVH sequelae). A prenatal origin can be assumed in term-born children with CP showing this injury, which is supported by the fact that these children usually have an unremarkable peri- and neonatal history. Cortical or deep grey matter lesions represent the second most frequent pattern in CP (around 20%, excluding ataxic CP). They were the typical lesions of term-born children with athetoid CP or severe BS-CP ( tetraplegia or quadriplegia ), reported in 44% of these children. Basal ganglia/thalamus or bilateral cortico-subcortical lesions suggest a peri- or neonatal origin, as these patterns are highly associated with peri- and neonatal compromise. 19 It has to be emphasized, however, that this conclusion cannot be drawn without a clear periand neonatal history supporting this assumption, ideally with sequential imaging (e.g. ultrasound) to document brain oedema or an evolving lesion. In the case of an uneventful peri- and neonatal period, a prenatal origin has to be assumed. This has been reported for basal ganglia/thalamus lesions in 23% of children in a hospital-based study. 54 In US- CP, cortical or deep grey matter lesions occurring in around a third of cases were mainly infarcts, and their precise timing is even more difficult to ascertain; convulsions during the neonatal period may indicate a neonatal origin, but neonatal or hypoxic ischemic encephalopathy is rare. 55,19 When cortical or deep grey matter lesions were reported in preterm-born children with CP, they tended to be of higher gestational age Developmental Medicine & Child Neurology 2007, 49:

6 Brain maldevelopments, indicating 1st and 2nd trimester origin, were rather rare, reported in less than 10% of children with CP. They occurred six times more often in termthan in preterm-born children with CP. The CP-subtype with a somewhat higher proportion of these patterns was US-CP (spastic hemiplegia), where cortical dysplasias, unilateral schizencephaly, and other focal malformations accounted for 16% of cases, and also ataxic CP where 17% of cases were reported to have cerebellar hypoplasias. A crucial question is whether the MRI abnormalities reported are responsible for the clinical picture of the CP. This issue has not been systematically addressed in most of the studies. The question of specificity could be addressed by studying MRI in a control group of normally developing children. Studies performed in normal developing children usually do not find incidental structural MRI abnormalities of cerebral parenchyma in more than 1 to 2% of participants. 56,57 Kim et al reported in more detail on 21% of incidental findings in a study of 225 neurologically healthy children, which were, however, mostly extracerebral or did not affect the cerebral parenchyma. In five children, they found changes that could be discussed within the MRI patterns above (three focal white matter lesions of uncertain aetiology, one ventricular asymmetry, and one cerebellar tonsil lesion of uncertain aetiology). These changes, obviously, did not interfere with neurological function. Thus, it seems important to study a lesion detected by MRI with respect to its topography and extent within the motor systems, and its relationship to Table II: Papers in Group II fulfilling part of inclusion criteria including 1022 magnetic resonance imaging (MRI) results from all cerebral palsy types; MRI described as abnormal in 930 (91%) Paper Number of Age at MRI Age at last Abnormal CP subtypes (n) Primary inclusion criteria participants follow-up, y MRI a not fulfilled because of: Truwit et al mo 41y > BS-CP (21) Early MRI Hypotonic (4) CP diagnosis unclear US-CP (5) (hypotonic CP) Choreoathetotic (1) Candy et al mo >2 17 Spastic diplegia (3) Early MRI Spastic quadriplegia (1) (7 children <1y at MRI) US-CP (2) CP definition unclear Rigid (3) (7 hypotonic CP) Choreoathetoid (2) Hypotonic (7), Mixed (4) Steinlin et al mo? 30 US-CP (30) Early MRI (18 children <1y at MRI) Niemann et al y >8 33 US-CP (33) MCA infarcts excluded Sugimoto et al mo 15y > 2 70 Tetraplegia (38) Early MRI US-CP (14) Diplegia (10) Ataxic diplegia (5) Dyskinetic CP (3) Hayakawa et al mo 13y? 56 Spastic diplegia (56) Early MRI Hayakawa et al mo 14y? 34 Spastic tetraplegia (34) Early MRI (17 <1y, 4 PVL <1y) Cioni et al a 8 36mo >2 48 Spastic diplegia (29) Early MRI, MR-pattern classification not Spastic quadriplegia (19) possible, maldevelopments excluded Steinlin et al /25 MRI 3 29y Non-progressive ataxia (10) Restricted to cerebellar hypoplasia Jaw et al mo 13y? 82 Spastic diplegia (20) Early MRI Spastic quadriplegia (20) CP definition unclear Hemiplegia (27) (atonic CP, unclassified CP) Monoplegia (2) Atonic (6), Athetoid (1) Mixed (3), Unclassified (7 ) Cioni et al a 7mo 13y 4mo >1 81 a US-CP (81) Early MRI Yin et al mo 18y? 39 Spastic diplegia (8) Early MRI Quadriplegia 20 (21) US-CP (10) 3 ataxia (1) Kulak et al y >4 83 Spastic diplegia (38) MR-pattern classification not possible Spastic tetraplegia (48) (cerebral atrophy in 17) Bax et al /351 MRI 1 87mo 12 91mo 310 US-CP (113) No distinction into term/preterm, Spastic diplegia (148) Subtype distribution in entire group Spastic quadriplegia (80) not MRI-specific Dyskinesia (62) ataxia (17) other (11) a Postneonatal CP excluded; BS-CP, bilateral spastic cerebral palsy; US-CP, unilateral spastic cerebral palsy; MCA, middle cerebral artery;?, unknown. Review 149

7 neurological signs and functional severity, ideally in comparison to a control group. The premise is that lesions within the pyramidal system (motor cortex or pyramidal tracts) account for spastic CP and lesions of the basal ganglia/thalamus for dyskinetic CP, and that the extent of the lesions within these systems relate to functional severity. 54,58 60 Yokochi et al. 28 and Krägeloh-Mann 25 looked at the extent of periventricular lesions in relation to severity of motor deficits and found that extensive or global ventricular enlargement, e.g. periventricular tissue loss, was accompanied by more severe functional deficit in BS-CP. Yokochi et al. 29 found that in athetoid CP, motor function was more severely affected in basal ganglia/thalamus lesions when they were accompanied by additional lesions than in deep grey matter or white matter lesions only. In conclusion, MRI has a high potential to elucidate aetiology, or at least pathogenesis, in CP. It confirms a mainly acquired, e.g. lesional origin in the majority of children with spastic and dyskinetic CP subtypes, which account for 94% of CP. 26 Ataxic CP is different; abnormal MRI findings occur in less than 40% and are, at least in half of the cases, not clearly lesional in the sense discussed above. There seems to be a consensus concerning the importance of MRI in the diagnostic work-up of CP.5 A consensus on the classification of MRI results is not yet established, however, and would be helpful also for CP registers which collect more and more data, also on imaging. We also think that the relation between structure and topography of the brain lesions and clinical function in children with CP merit further investigation as they are important prerequisites for questions of reorganization and plasticity. Accepted for publication 23rd November Acknowledgement We are grateful to Prof. Philippe Evrard. Paris, for the discussion on the pathogenesis of early brain lesions. References 1. Stanley F, Blair E, Alberman E. (2000) Cerebral palsies: Epidemiology and Causal Pathways. Clinics in Developmental Medicine No London: Mac Keith Press. 2. Himmelmann K, Hagberg G, Beckung E, Hagberg B, Uvebrant P. (2005) The changing panorama of cerebral palsy in Sweden IX. Prevalence and origin in the birth-year period Acta Paediatr 94: Bax M, Goldstein M, Rosenbaum P, Leviton A, Paneth B, Dan B, Jacobsson B, Damiano D. (2005) Proposed definition and classification of cerebral palsy. Dev Med Child Neurol 47: SCPE working group. (2000) Surveillance of cerebral palsy in Europe: a collaboration of cerebral palsy surveys and registers. Dev Med Child Neurol 42: Ashwal S, Russman BS, Blasco BA, Miller G, Sandler A, Shevell M, Stevenson R. (2004) Practice Parameter: Diagnostic assessment of the child with cerebral palsy. Neurology 62: Barkovich J. (2000) Pediatric Neuroimaging, 3rd ed. Raven Press. 7. Evrard PH, Miladi N, Bonnier CH, Gressens P. (1992) Normal and abnormal development of the brain. In: Rapin I, Segalowitz SJ, editors. Handbook of Neuropsychology 6: Child Neuropsychology , Eds Rapin I, SJ Segalowitz, Amsterdam: Elsevier. p Barkovich AJ, Kuzniecky RI, Jackson GD, Guerrini R, Dobyns WB. (2001) Classification system for malformations of cortical development. Neurology 57: Johnston MV, Trescher WH, Ihida A, Nakajima W. (2001) Neurobiology of hypoxic-ischemic injury in the developing brain. Pediatric Research 49: Hagberg H, Mallard C. (2005) Effects of inflammation on central nervous system development and vulnerability. Curr Opin Neurol 18: Edwards AD, Tan S. (2006) Perinatal infections, prematurity and brain injury. Current Opinions in Pediatrics 18: Dommergues MA, Palkai J, Renauld JC, Evrard P, Gressens P. (2000) Proinflammatory cytokines and IL-9 exacerbate excitotoxic lesions in the newborn murine neopallidum. Ann Neurol 47: Hagberg H, Peebles D, Mallard C. (2002). Models of white matter injury: comparison of infectious, hypoxic-ischemic and excitotoxic insults. Ment Retard Dev Disabil Res Rev 8: Keeney SE, Adcock EW, McArdle CB. (1991) Prospective observations of 100 high-risk neonates by high-field (1.5 Tesla) magnetic resonance imaging of the central nervous system. II. Lesions associated with hypoxic-ischemic encephalopathy. Pediatrics 87: Baenziger O, Martin E, Steinlin M, Good M, Largo R, Burger R, Fanconi S, Duc G, Buchli R, Rumpel H, Boltshauser E. (1993) Early pattern recognition in severe perinatal asphyxia: a prospective MRI study. Neuroradiology 35: Rutherford MA, Pennock JM, Schwieso JE, Cowan FM, Dubowitz LMS. (1995) Hypoxic ischaemic encephalopathy: Early magnetic resonance imaging findings and their evolution. Neuropediatr 26: De Vries LS. (1996). Neurological assessment of the preterm infant. Acta Paediatrica 85: Volpe JJ. (1997) Brain injury in the premature infant from pathogenesis to prevention. Brain & Development 19: Cowan F, Rutherford M, Groenendaal F, Eken P, Mercuri E, Bydder GM, Meiners LC, Dubowitz LMS, de Vries LS. (2003) Origin and timing of brain lesions in term infants with neonatal encephalopathy. The Lancet 361: Hagberg B, Hagberg, G, Olow I, L.v.Wendt. (1996) The changing panorama of cerebral palsy in Sweden. VII. Prevalence and origin in birth year period Acta Paediatr 85: Govaert P, Matthys E, Zecic A, Roelens F, Oostra A, Vanzieleghem B. (2000) Perinatal cortical infarction within middle cerebral artery trunks. Arch Dis Child Fetal Neonatal 82: F59 F De Vries LS, Groenendaal F, Eken P, van Haastert IC, Rademaker KJ, Meiners LC. (1997). Infarcts in the vascular distribution of the middle cerebral artery in preterm and fullterm infants. Neuropediatr 28: Michaelis R, Asenbauer C, Buchwald-Saal M, Haas G, Krägeloh- Mann I. (1993) Transitory neurological findings in a population of high-risk infants. Early Human Development 43: Dehan M, Vodovar M, Goujard J, Crost M, Rougeot M, Gautier JP, Benisvy C, Plissier M, Voyer M. (1990) Premature deliveries before the 33rd week of gestational age: an enquiry carried out in 1985 in the Paris (geographical) region. J Gynecol Obstet Biol Reprod 19: Krägeloh-Mann I. (2004) Imaging of early brain injury and cortical plasticity. Exper Neurol 190: Surveillance of Cerebral Palsy in Europe (SCPE) working group. (2002) Prevalence and characteristics of children with cerebral palsy in Europe. Dev Med Child Neurol 44: Koeda T, Suganuma I, Takamatsu T, Takeshita K. (1990) MR imaging of spastic diplegia. Comparative study between preterm and term infants. Neuroradiology 32: Yokochi K, Aiba K, Horie M, Inukai K, Fujimoto S, Kodama M, Kodama K. (1991) Magnetic resonance imaging in children with spastic diplegia: Correlation with the severity of their motor and mental abnormality. Dev Med Child Neurol 33: Yokochi K, Aiba K, Kodama M, Fujimoto S. (1991) Magnetic resonance imaging in Athetotic cerebral palsied children. Acta Paediatr Scand 80: Krägeloh-Mann I, Petersen D, Hagberg G, Vollmer B, Hagberg B, Michaelis R. (1995) Bilateral spastic cerebral palsy MRI pathology and origin. Analysis from a representative series of 56 cases. Dev Med Child Neurol 37: Okumura A, Kato T, Kuno K, Hayakawa F, Watanabe K. (1997) MRI findings in patients with spastic cerebral palsy. ll: correlation with type of cerebral palsy. Dev Med Child Neurol 39: Kwong KL, Wong YC, Fong CM, Wong SN, Kwan TS. (2004). Magnetic resonance imaging in 122 children with spastic cerebral palsy. Pediatric Neurol 31: Developmental Medicine & Child Neurology 2007, 49:

8 33. Esscher E, Flodmark O, Hagberg G, Hagberg B (1996) Nonprogressive ataxia: origins, brain pathology and impairments in 78 Swedish children. Dev Med Child Neurol 38: Truwit ChL, Barkovich AJ, Koch TK, Ferriero DM. (1992) Cerebral palsy: MR findings in 40 patients. AJNR 13: Candy EJ, Hoon AH, Capute AJ, Bryan RN. (1993) MRI in motor delay: important adjunct to classification of cerebral palsy. Pediatric Neurol 9: Steinlin M, Good M, Martin E, Bänsiger O, Largo RH, Boltshauer E. (1993) Congenital hemiplegia: morphology of cerebral lesions and pathogenetic aspects from MRI. Neuropediatr 24: Niemann G, Wakat J-P, Krägeloh-Mann I, Grodd W, Michaelis R. (1994) Congenital hemiparesis and periventricular leukomalacia: pathogenetic aspects on magnetic resonance imaging. Dev Med Child Neurol 36: Sugimoto T, Woo M, Nishida N, Araki A, Hara T, Yasuhara A, Kobayashi Y, Yamanouchi Y. (1995) When do brain abnormalities in cerebral palsy occur? An MRI study. Dev Med Child Neurol 37: Hayakawa K, Kanda T, Hashimoto K, Okuno Y, Yamori Y, Yuge M, Ando R, Ozaki N, Tamamoto A. (1996) MR imaging of spastic diplegia. The importance of corpus callosum. Acta Radiologica 37: Hayakawa K, Kanda T, Hashimoto K, Okuno Y, Yamori Y. (1997) MR of spastic tetraplegia. AJNR 18: Cioni G, Di Paco MC, Bertucelli B, Paolicelli PB, Canapicchi R. (1997) MRI findings and sensorimotor development in infants with bilateral spastic cerebral palsy. Brain & Development 19: Steinlin M, Zanegger B, Bolthauser E. (1998) Non-progressive ataxia with or without cerebellar hypoplasia: a review of 34 subjects. Dev Med Child Neurol 40: Jaw, T-S, Jong Y-J, Sheu R-S, Liu G-C, Chou M-S, Yang R-C. (1998) Eetiology, timing of insult, and neuropathology of cerebral palsy evaluated with magnetic resonance imaging. J Formos Med Assoc 97: Cioni G, Sales B, Paolicelli PB, Petacchi E, Scusa MF, Canapicchi R. (1999) MRI and clinical characteristics of children with hemiplegic verebral palsy. Neuropediatrics 30: Yin R, Reddihough DS, Ditchfield MR, Collins KJ. (2000) Magnetic resonance imaging findings in cerebral palsy. J Paediatr Child Health 36: Kulak W, Sobaniec W, Smiglielska-Kuzia J, Kubas B, Walecki J. (2005) A comparison of spastic diplegic and tetraplegic cerebral palsy. Pediatric Neurol 32: Bax, M, Tydeman C, Flodmark O. (2006) Clinical and MRI correlates of cerebral palsy. The European cerebral palsy study. Jama 296: Doyle LW, Betheras FR, Ford GW, Davis NM, Callanan C. (2000). Survival, cranial ultrasound and cerebral palsy in very low birthweight infants: 1980s versus 1990s. J Paediatr Child Health 36: Topp M, Uldall P, Greisen G. (2001) Cerebral palsy births in eastern Denmark, : implications for neonatal care. Paediatr Perinat Epidemiol 15: Surman G, Newdick H, Johnson A. (2003) Cerebral Palsy rates among low-birthweight infants fell in the 1990s. Dev Med Child Neurol 45: Platt MJ, Cans C, Johnson A, Surman G, Topp M, Torrioli MG, Krägeloh-Mann I. Trends in cerebral palsy among infants of very low birthweight (<1500g) or born prematurely (<32 weeks) in 16 European centers. Lancet. (Forthcoming) 52. Watson L, Stanley F, Blair E. (1999) Report of the Western Australia CP Register. Perth: Western Austrailia CP Register. 53. Winter S, Autry A, Boyle C, Yeargin-Allsopp M. (2002) Trends in the prevalence of cerebral palsy in a population-based study. Pediatrics 110: Krägeloh-Mann I, Helber A, Mader I, Staudt M, Wolff M, Groenendaal F, DeVries L. (2002) Bilateral lesions of thalamus and basal ganglia: origin and outcome. Dev Med Child Neurol 44: Levy SR, Abrams IF, Marshall PC, Rosquette EE. (1985) Seizures and cerebral infarction in the fullterm newborn. Ann Neurol 17: Byars AW, Holland SK, Strawsburg RH, Schmithorst VJ, Dunn RS, Ball WS. (2002) Practical aspects of conducting large-scale fmri studies in children. J Child Neurol 17: Kim BS, Illes J, Kaplan RT, Reiss A, Atlas SW. (2002) Incidental findings on pediatric MR images of the brain. American J Neuroradiol 23: Staudt M, Niemann G, Grodd W, Krägeloh-Mann I. (2000) The pyramidal tract in congenital hemiparesis: relationship between morphology and function in periventricular lesions. Neuropediatrics 31: Staudt M, Pavlova M, Böhm S, Grodd W, Krägeloh-Mann I. (2003) Pyramidal tract damage correlates with motor dysfunction in bilateral periventricular leukomalacia (PVL). Neuropediatrics 34: Thomas B, Eyssen M, Peeters R, Molenaers G, Van Hecke P, De Cock P, Sunaert S. (2005) Quantitative diffusion tensor imaging in cerebral palsy due to periventricular white matter injury. Brain 128: Review 151