A cognitive and affective pattern in posterior fossa strokes in children: a case series


Dr Manoelle Kossorotoff at Service de Neurologie pédiatrique, Hôpital Necker-Enfants Malades, 149 rue de Sèvres, 75015 Paris, France. E-mail: manoelle.kossorotoff@nck.aphp.fr


Aim  Posterior fossa strokes account for about 10% of ischaemic strokes in children. Although motor and dysautonomic symptoms are common, to our knowledge cognitive and affective deficits have not been described in the paediatric literature. Our aim, therefore, was to describe these symptoms and deficits.

Method  In a retrospective study, we included all cases of posterior fossa strokes in children occurring at a single centre between 2005 and 2007, and investigated cognitive and affective deficits.

Results  Five males aged 3 to 14 years met the inclusion criteria. They all presented very early with mood disturbances: outbursts of laughter and/or crying and alternating agitation or prostration that disappeared spontaneously within a few days. Persistent cognitive deficits were also diagnosed in all five: initial mutism, then anomia, followed by comprehension deficiency and deficiencies of planning ability, visual–spatial organization, and attention. Despite early and intensive rehabilitation, recovery from these cognitive deficits was slow and sometimes incomplete, and on follow-up they proved to be more disabling than the motor symptoms.

Interpretation  These findings are similar to the cerebellar cognitive affective syndrome described in adults, and quite similar to the language and affective deficits observed in children after surgery for posterior fossa tumour. This is consistent with the role of the cerebellum and brainstem in affective and cognitive processes from early development.

The incidence of acute ischaemic stroke in children (excluding the neonatal period) is estimated to be 1.2 per 100 000 per year,1–4 and it is estimated that posterior fossa stroke (PFS) accounts for 5 to 15% of acute ischaemic strokes in children.5–7

Cerebellar cognitive and affective syndrome has been reported in adults with PFS, who present with cognitive deficits that include executive dysfunction (perseveration, lack of planning, deficit of lexical retrieval and of working memory), visual–spatial disorganization and inattention, visual–spatial memory deficiency, language difficulties (dysprosodia, agrammatism, word-finding difficulties), and affective signs with altered personality (apathy, disinhibition, inappropriate behaviour).8 A degree of correlation between the type of cognitive deficits and lesion topography has been reported.8 The identification of this syndrome in adults is further evidence of cerebellar involvement in higher functions, including language comprehension and production (syntax, lexical evocation), and visual–spatial, behavioural, and executive functions.

Similar cognitive and affective deficits have been reported in children after surgery for posterior fossa tumour; they have not, so far, been reported in children with PFS. PFS is usually considered uncommon in children; the main symptoms are said to be motor, sensory, and dysautonomic, while affective symptoms are rarely mentioned. We recorded these signs and symptoms in a series of children with PFS. The aim of this paper is to describe these deficits in children with this disorder.


We reviewed the charts of all children admitted to the child neurology unit of Necker-Enfants Malades University Hospital in Paris between 2005 and 2007 to identify those affected by stroke, excluding neonatal strokes. We included in the study all children exhibiting clinical and radiological evidence of ischaemic PFS occurring after the age of 28 days and before the age of 18 years. Five children met the inclusion criteria. In each case, we analysed medical history, age at onset, type of stroke, localization, and characteristics of the stroke (focusing on cognitive and affective symptoms and signs), together with the type of rehabilitation and quality of recovery. All five children underwent early brain magnetic resonance imaging (MRI), including T2-weighted, fluid-attenuated inversion recovery (FLAIR), and diffusion-weighted imaging (DWI) sequences, and apparent diffusion coefficient mapping (before day 7 after the first signs of stroke).

Brain MRI was again performed, and a neuropsychological assessment carried out, more than 6 months after the acute stroke. Neuropsychological tests were selected according to the national French protocol, ‘Study of the effects of central nervous system hyperfraction radiotherapy for the treatment of medulloblastoma in children’ (Institut Gustave Roussy).9 Our comprehensive neuropsychological battery10,11 included (1) verbal and non-verbal intellectual functioning evaluation using Wechsler’s scale (WISC-III); (2) visual–spatial testing with the Rey’s complex figure copy; (3) memory testing with Rey’s complex figure recall and Rey’s auditory learning test; (4) language evaluation using a vocabulary test (the Peabody Picture Vocabulary Test – Revised (EVIP in French), a naming test (DEN 48): to verbally identify 48 line drawings of real objects), verbal fluency tasks (1min words generation: animal names and words beginning with the letter ‘m’), and reading tasks testing speed (reading words loudly) and comprehension (Khomsi’s LMC-R, Lefavrais); and (5) executive functions evaluation by the Wisconsin Card Sorting Test (WCST) and the Stroop test.


All five participants were males aged between 3 years 6 months and 14 years. Four of them had normal psychomotor development (including one with developmental dyslexia) and one (participant 4) had mild difficulties at school. Four had no medical history, but one had previously undergone vascular surgery for congenital aortic coarctation at birth and at 2 and 6 years of age. In all cases, early clinical and radiological data as well as clinical data and the findings of neuropsychological assessment and brain MRI 6 months after stroke onset were available. Participant characteristics are provided in Table I.

Table I.   Participant descriptions – stroke characteristics, observed signs and symptoms at the early stage, results of cognitive assessment more than 6 months after the stroke
 Signs at early stageNeuropsychological assessment >6 m after PFS)
Participant no.Sex and age at first strokeVascular involvementStroke locationInitial motor signsDysautonomic signsCognition (except language)Language and speech deficitsAffective signsIQ (WISC-III/IV)Rey’s drawing: copy (C), recall (R)Fluency: animals (A) letter ‘m’ (L)Naming (DEN 48)Vocabulary (EVIP)Reading comprehension (Khomsi’s, LMC-R, Lefavrais) (C), reading speed (S)
  1. aSee ‘Results, late stage’ section in the text for further description of the second stroke. DEN 48, a naming test to verbally identify 48 line drawings of real objects; EVIP, Peabody Picture Vocabulary Test – Revised (in French); NP, not performed; perc., percentile; PIQ, Performance IQ; VIQ, Verbal IQ; WISC, Wechsler Intelligence Scale for Children.

1Male, 14y 10moRight superior cerebellar artery embolism (endocarditis, aortic coarctation, repeated cardiac surgery)Right cerebellar hemisphere, vermis and right brainstemRight cerebellar static and kinetic syndrome, right facial palsy, right VI nerve palsy, left hemiparesisNoneVisual–spatial disorganization, planning difficultiesLanguage production disorders: mild anomia, decreased fluency, followed by dysarthriaFlattened and negative affects, hypersomnia, catatonia (<7d)VIQ 113, PIQ 90C +0.5SD, R +3.2SDA −0.8SD, L −2.2SD+0.7SDperc. 55C <perc.10, S perc. 75
2Male, 8yLeft vertebral artery dissection and basilar artery embolism (partial)Right cerebellar hemisphere and vermis (first episode), brainstem bilaterally (second episode)aRight cerebellar kinetic syndrome, left hemiparesis, VI and XII nerve palsyCephalalgia, repeat vomiting, then blood pressure variability and bradycardia, bruxism, yawningVisual–spatial disorganization, executive dysfunction, attention, working memory, and planning difficultiesMutism followed by language production disordersAnxiety, agitation, mood instability, disinhibition, outbursts of laughter and/or crying (5d)VIQ 85, PIQ 66Within normal rangeA −1SD, L −0.7SD−0.7SDperc. 64C perc. 50, S perc. 10
3Male, 9y 8moBasilar artery occlusionPons, predominantly rightDysarthria, left hemiparesis, Left facial palsySevere sinus bradycardiaAttention lability, memory deficitMutism followed by language production disorders (anomia, slow speech, dysarthria), difficulties in reading comprehensionAgitation, outbursts of laughter and/or crying (3d)VIQ 105, PIQ 61C −3.5SD, R −1SDA −0.4SD, L −0.5SD−4.5SDperc. 85C perc. 50–75, S perc. 50
4Male, 9y 8moBasilar artery occlusionPons, predominantly rightLeft hemiparesis, then hemiplegia, left facial palsy, left XII nerve palsyCephalalgia, mild sinus bradycardiaPlanning difficulties, Memory difficultiesLanguage production disorders: dysarthria, syntactic deficiencyOutbursts of laughter and/or crying (4d)VIQ 53, PIQ 99C −1.2SD, R −4.6SDA +0.8SD, L test failureNPperc. 6C <perc. 5, S <perc. 10
5Male 3y 5moVertebral artery dissection and basilar artery embolism (distal occlusion)Pons, predominantly leftRight hemiparesis, right facial palsy, horizontal nystagmus, swallowing difficultiesCephalalgia, repeat vomiting, bradyarrhythmiaAttention labilityMutism followed by dysarthriaOppositionVIQ 128, PIQ 94C +0.9SD, R +1.6SDA +1.3SD, L −0.5SD−0.6SDNPC <perc. 10


Early stage (during first week)

Stroke presented in a multi-step fashion in four out of the five participants (participants 2–5); the time elapsing between the first symptoms and complete stroke ranged from 1 hour to 10 days. The initial signs were diverse: cephalalgia or cervical pain (4/5); drowsiness and repeated vomiting (2/5), in one case preceded by sudden and brief loss of consciousness (participant 2); isolated dysarthria (participant 3); and cerebellar motor signs (participant 1). Once stroke was complete, all five participants exhibited motor signs: pyramidal signs with hemiparesis (5/5), cerebellar motor signs (2/5; kinetic dysmetria in both participants and static ataxia in one), cranial nerve deficits (5/5), and dysarthria (4/5). Four out of five children presented with autonomic signs, including mild to severe sinus bradycardia or bradyarrhythmia.

All participants presented with homogeneous affective features characterized by mood disorders: alternating inappropriate crying and laughing (3/5), agitation and opposition (3/5), and sad mood (1/5). These signs occurred very early, before arrival at hospital, and they disappeared spontaneously in less than a week. Expressive language difficulties were also noted in all five participants, ranging from decreased verbal fluency and mild anomia to mutism. Brain MRI confirmed acute ischaemic infarction involving the brainstem in all five participants and the cerebellar vermis and hemisphere in two participants. All the participants underwent multidisciplinary rehabilitation: physiotherapy (5/5), splint (4/5), occupational therapy (4/5), speech therapy (5/5), and, in the case of participant 5, psychomotility therapy and psychotherapy.

Late stage

Neuropsychological testing performed after more than 6 months revealed cognitive deficiencies in all cases. A uniform and specific pattern was delineated in four participants. The results in participant 4 were more difficult to analyse because of previous language and schooling difficulties. Although global cognitive level was within the normal range in all cases, all participants except participant 4 presented with a dissociated cognitive profile, with better verbal than non-verbal skills (Verbal IQ>Performance IQ+12). The striking expressive language difficulties initially observed during the acute phase (mutism, decreased verbal fluency, mild anomia) had at this time decreased in intensity and changed into less obvious slow speech, word-finding difficulties, and agrammatism and dysprosodia. Three children exhibited reading comprehension difficulties.

In terms of non-verbal skills, participants exhibited attention lability (4/5), planning difficulties and visual–spatial disorganization (2/5), and visual–spatial memory deficiency (3/5). Three participants exhibited working memory deficiency (3/5) observed through the memory span in the Wechsler’s scale. Four out of five participants were receiving normal schooling in spite of these various cognitive difficulties and persistent mild speech disorders. Regarding behaviour and affective deficits, the participants’ families reported higher than average anxiety and difficulty in concentrating. During the tests, children were cooperative but slow, and they had poor self-esteem.

Neuropsychological findings are shown in Table I. One participant (participant 2) experienced a second stroke 3 months after the first one, presenting with a left pyramidal syndrome, left eye deviation, aphasia, problems with consciousness, bradycardia, and yawning. Brain MRI revealed massive brainstem involvement, leading to the diagnosis of vertebral dissection with partial basilar artery occlusion. Brain MRI more than 6 months after the stroke revealed mild atrophy with gliosis involving the brainstem (5/5) and cerebellum (2/5) (Figure 1). Physical recovery was relatively good in all instances, with a follow-up ranging from 6 months to 8 years. All five children had mild to moderate unilateral motor sequelae (hemiparesis in five and persisting dysmetria in two), but were able to walk and write, although they exhibited mild writing difficulties (5/5).

Figure 1.

 Early-stage brain magnetic resonance images in the axial plane from participants 1–4. (a) Participant 1: axial T2 image 2 days after the onset of stroke. The solid arrow shows right cerebellar infarction. (b) Participant 2: axial T2 image taken 1 day after the onset of the first stroke. The solid arrow shows right cerebellar infarction. (c) Participant 2: axial T2 image taken 3 days after the onset of the second stroke. The dashed arrow shows the sequelae of the first stroke, whereas the solid arrow shows recent brainstem infarction. (d) Participant 3: axial fluid-attenuated inversion recovery (FLAIR) image 3 days after the onset of the stroke. The solid arrow shows the bilateral pons infarction. (e) Participant 4: axial T2 image taken 2 days after the onset of stroke. The solid arrow shows the predominantly right pons infarction.


In our series, PFS affected five out of 51 children with strokes (excluding neonatal strokes) during the period 2005 to 2007. This is consistent with the available epidemiological data.1,2,4

The course was unexpected, with recovery being surprisingly better than that reported in adults with PFS. This favourable outcome was all the more unpredictable because signs of severity (autonomic signs and signs of lower cranial nerve involvement) were present in four of the participants at the early stage. Three of these five participants presented with basilar artery occlusion and made a good recovery, going on to normal schooling, despite the fact that basilar artery occlusion is known to have a fairly poor prognosis.12 Although mood and cognitive disorders are rarely reported in paediatric PFS,13 all five participants initially presented with a specific affective and cognitive profile. The features in our paediatric participants are quite similar to the cerebellar cognitive affective syndrome reported in adult participants.8,13 However, affective symptoms were transient, persisting for a shorter time than motor or cognitive signs, and they could have easily been overlooked had the participants’ cognitive skills and behaviour not been examined.

Schmahmann and Sherman8 first described this clinical pattern in 20 adults with various congenital or acquired cerebellar diseases, including 13 with stroke.8,13 The cognitive and affective dysfunction was attributed to cerebellar dysfunction. Reported cognitive deficits were difficulties in language production, including decreased fluency, mild anomia, agrammatism, and dysprosody. Impairment of executive functions was associated with planning difficulties, visual–spatial disorganization, and impaired visual–spatial and working memory. Attention lability was also reported. Affective symptoms consisted of a change of personality, with blunting of affect or disinhibited and inappropriate behaviour. A similar cognitive and affective profile was later reported in cerebellar strokes.13–15

A similar clinical pattern was reported in children with posterior fossa tumour.16 Cerebellar mutism after posterior fossa surgery was described in the 1990s,17,18 followed by more complete descriptions of cognitive and affective symptoms.19 Irradiation was suspected, but children treated with surgery alone also present with postoperative neuropsychological deficits.16,19,20 In 26 children who had undergone surgery for the removal of cerebellar tumour, Riva and Giorgi21 reported different patterns according to tumour location. Children with right cerebellar tumour presented with auditory sequential memory and language processing disturbances, whereas those with left cerebellar tumour showed deficits in spatial and visual sequential memory. Children with vermal lesions exhibited postsurgical mutism followed by speech disturbances (agrammatism), or behavioural disturbances ranging from irritability to autistic behaviour.21 Mood symptoms were described, with bizarre behaviour, emotional lability, extreme irritability, and inhibition.22 According to several series, lesions of the vermis or the floor of the fourth ventricle are associated with major affective alterations, whereas those involving cerebellar hemispheres result in altered cognitive abilities.9,16,21,23

There are differences between our participants and the children described in postoperative series. We did not find any clear relationship between lesion localization and the type of cognitive and behaviour symptoms. All children presented with similar affective symptoms (agitation and mood disorders, with bouts of laughter and crying), although lesions affected the cerebellum in only two participants. Nor was the type of cognitive disturbance determined by the localization of stroke: all five participants exhibited homogeneous language disturbances, mainly affecting productive domains, as reported in vermal tumours, although only two had vermal lesions. The common lesion site shared by all our participants was the brainstem, suggesting that injury to the cerebellar–cerebral pathways leads to disorders similar to those produced by cerebellar injuries, supporting a diaschisis-like mechanism.

Another important difference is that the mood disorder in our participants was less severe than that which occurs after surgery, and was reversible within less than a week, whereas mood disorders in postoperative participants persist for several years. Neuropsychological deficits after stroke were milder in the long term than reported postoperatively.9 The early occurrence and brief course of affective signs point to a transient and reversible phenomenon, suggesting impairment caused by oedema, whereas other signs (motor and cognitive) persisted, reflecting a destructive mechanism.

These cognitive and affective deficits can be compared with the cortical dysfunction described in postsurgical PFS in children. Miller et al.24 suggested that in these children the pathophysiology is surgical disruption of cerebellar–cerebral connections, in a diaschisis-like mechanism. In fact, the diaschisis phenomenon was initially described in non-surgical conditions, such as supratentorial stroke, with hypometabolism and hypoperfusion in the contralateral cerebellar hemisphere. Diaschisis is classically defined as a sudden inhibition of function in an area of the brain remote from, but anatomically connected to, the site of the primary injury. We propose a similar mechanism in our participants. In terms of implicated neuronal circuits, reciprocal connections between the cerebrum and cerebellum mainly involve frontal structures, cingulate, and hypothalamus. Outbursts of laughter and crying may suggest frontal disinhibition. The existence of cognitive affective syndrome in children with PFS supports a role for the cerebellum and other infratentorial structures in higher cognitive function and mood control from a very early age.21

Our study was retrospective, and some short-lasting signs may not have been recognized. We plan to evaluate mood disorders at an early stage using specific items, such as the Achenbach’s Child Behaviour Checklist. It would be interesting in a future study to perform MRI fibre tracking in children with PFS to determine support for the hypothesis that a diaschisis-like mechanism plays a role in the pathophysiology of cognitive and affective deficits. Further, to evaluate the impact of acquired versus congenital cerebellar lesions, we plan to compare the cognitive and affective profiles of children with PFS and those who have experienced perinatal cerebellar stroke.


Early transient affective and persistent cognitive dysfunction was observed in five children with PFS, in association with cerebellar motor symptoms. Affective signs were transient and fully reversible. These deficits should be recognized as neurological manifestations, not merely as the child’s reaction to the motor deficit. Our study supports the existence of a paediatric ‘posterior fossa cognitive affective syndrome’ after stroke.

What this paper adds

  • • Our findings support the existence of a paediatric ‘posterior fossa cognitive affective syndrome’ after stroke, as described postoperatively in children with posterior fossa tumour.
  • • Affective symptoms are precocious and transient.
  • • Cognitive deficits should be recognized because they are disabling in the long term.
  • • The pathophysiology of posterior fossa strokes could involve the diaschisis phenomenon.


M Kossorotoff and C Gonin-Flambois contributed equally to this research.