OUP user menu

Worster-Drought syndrome, a mild tetraplegic perisylvian cerebral palsy
Review of 47 cases

Maria Clark, Lucinda Carr, Sheena Reilly*, Brian G. R. Neville
DOI: http://dx.doi.org/10.1093/brain/123.10.2160 2160-2170 First published online: 1 October 2000


A retrospective case-note analysis was undertaken of 47 children with a congenital upper motor neurone bulbar palsy (excluding pure speech dyspraxia) to clarify the phenotype of Worster-Drought syndrome (WDS) and to record its associated features and complications. The results revealed that the study children had significant bulbar problems (with 80% still needing a modified diet and a similar number using augmentative communication methods at last review). There were also high rates of predictable bulbar complications (86% had dribbling, 60% had glue ear, gastro-oesophageal reflux in 40%, history of poor nutrition in 40% and aspiration in 40%). Most of the children had additional complex impairments (91% had mild pyramidal tetraplegia, 81% learning difficulties, 60% congenital defects, 41% neuropsychiatric problems and 28% epilepsy). Over half of the children had significant medical problems in the first year, but mean age at diagnosis was 6 years. There were no obvious causes in pregnancy or birth. Six children had a family history of WDS and 32% (12/37) had abnormal neuroimaging including five with bilateral perisylvian polymicrogyria. In our experience, WDS is not uncommon, is relatively easily diagnosed and is crucial not to miss as the management of these children's multiple impairments is complex and requires a careful team approach. WDS falls clearly within the cerebral palsies as a syndrome that includes motor impairment arising from static damage to the brain in early life. The common presence of cognitive, behavioural and seizure impairments strongly supports the cerebral cortical (presumably perisylvian) localization. Its core elements are a suprabulbar paresis, a mild spastic tetraplegia and a significant excess of cognitive and behavioural impairments and epilepsy. The complete overlap in phenotype between WDS and the bilateral perisylvian syndrome leads us to propose that they are the same condition. WDS is startlingly absent from epidemiological studies of the cerebral palsies and rarely diagnosed, presumably because of lack of clinical awareness of the condition and lack of major gross motor impairments.

  • suprabulbar
  • tetraplegia
  • neuropsychiatric
  • epilepsy
  • perisylvian
  • CBPS = congenital bilateral perisylvian syndrome
  • WDS = Worster-Drought syndrome


The Worster-Drought syndrome (WDS) is a form of cerebral palsy originally described by Dr Worster-Drought, in which the major motor impairment is a pseudobulbar paresis (Worster-Drought, 1956, 1974). It appears to be poorly recognized; there are few descriptions of it in medical literature and no agreed diagnostic criteria. Indeed it has been omitted in most epidemiological studies of the cerebral palsies.

At Great Ormond Street Hospital, we see a large number of children with a congenital pseudobulbar palsy, and have found WDS a useful clinical diagnostic category. These children have a characteristic phenotype, easily recognized by paediatricians, therapists and parents, and are distinct within the cerebral palsies in being ambulant with severe expressive communication and feeding problems.

In our experience, an isolated pseudobulbar palsy appears comparatively rare. Most children have a mild four limb pyramidal tetraplegia with a variety of associated impairments (e.g. learning difficulties, behavioural problems, epilepsy and congenital defects), which tend to dominate management issues. The children appear to be diagnosed relatively late (average age at diagnosis of 6 years in this study) if at all. This delay may have long-term detrimental effects on function if it means there has not been an integrated approach to these children's multiple impairments. There appears to be a major degree of overlap with the condition described in the epilepsy literature as `bilateral perisylvian syndrome'.

The family support group has been a significant source of information and has a very clear position about the identity of this condition, from both their experience of their children's impairments and from the need for medical professionals to recognize their complex needs (Neville, 1997).

This retrospective study was devised to examine the following premises: (i) WDS has a distinctive phenotype; (ii) there are complex, but predictable management issues for these children which make early diagnosis important; and (iii) better delineation of the phenotype would further our understanding of pathogenesis.


Original description of Worster-Drought syndrome

Worster-Drought first described the syndrome in 1956 as congenital suprabulbar paresis (Worster-Drought, 1956, 1974). He considered that the primary lesion was congenital and situated in the cortico-bulbar tract, resulting in weakness or paralysis of the lips (orbicularis oris), tongue, soft palate, pharynx and laryngeal muscles, and an exaggerated jaw jerk. Clinical manifestations predominantly affected speech, and in severe cases there was usually constant dribbling, some dysphagia and occasionally even phonation problems.

Worster-Drought maintained that the syndrome was not difficult to diagnose. He stated that the child should always have a normal smile, but would have variable difficulty in moving the tongue, soft palate and lips, with dysarthric slurred nasal speech, dribbling and a brisk jaw jerk. The rest of their neurology, he considered, should be normal. He also asserted that regular speech therapy was needed in most cases (although he tempered this view in later writings) and that the majority (especially the milder cases) showed a gradual spontaneous improvement.

Worster-Drought excluded acquired cases from his definition, and felt that the acquired situation could be distinguished by history and by the fact that there would be additional neurological and behavioural findings. He also thought that developmental dyspraxia could be differentiated by the absence of any definite paresis affecting the lips, tongue or soft palate.

Lesional basis of pseudobulbar palsy

The neurological formulation developed by Worster-Drought proposed two possible localizations which were based on an adult acquired neurological conceptual framework, despite excluding acquired disorders. These were proposed as corticobulbar and cortical (although he favoured the corticobulbar tract). However, the congenital situation is likely to be more complex and less predictable than in the adult because of the developmental mechanisms of damage which may have widespread effects on the developing CNS.


In the adult, the corticobulbar fibres to the brainstem may be affected by occlusion of one of the lenticulostriate vessels from the middle cerebral artery causing a capsular infarction. Most of these brainstem nuclei have 50 : 50 innervation from each hemisphere so that unilateral lesions have little effect, whereas bilateral lesions (often sequential) produce pseudobulbar palsy and cause serious difficulty with speaking, chewing and swallowing, with little prospect of recovery. This syndrome is acquired and commonly associated with emotional lability.

Occasionally in this situation, the pseudobulbar palsy can be associated with an acquired oculomotor apraxia due to bilateral lesions affecting the descending pathways from the frontal eye field in the anterior internal capsule. The corticobulbar tracts are also characteristically involved in an insidious way in conditions such as progressive supranuclear palsy and some forms of motor neurone disease.

Worster-Drought proposed that in most cases, congenital pseudobulbar paresis was due to agenesis or hypogenesis of the corticobulbar tract.


There is also a cortical type of pseudobulbar palsy resulting from lesions in the opercula of both cerebral hemispheres. The operculum is the small region covering the anterior insula, and composed of contributions from the frontal, temporal and parietal gyri. This type of pseudobulbar palsy is characterized clinically by loss of voluntary functions of the muscles of the face, tongue, mastication and swallowing, with preservation of reflex and automatic functions. It is not associated with emotional lability.

The cortical form of pseudobulbar palsy is known as faciopharyngoglossomasticatory diplegia or Foix–Chavany–Marie syndrome, and is associated with clear bilateral perisylvian abnormalities on MRI scanning. In this situation, there is automatic–voluntary dissociation with loss of voluntary control of facial, pharyngeal, lingual and masticatory muscles, but preservation of reflexive and automatic functions. It is usually due to an acquired lesion (Moodley and Bamber, 1990), and has been described in a reversible form in children with epilepsy (Fusco and Vigevano, 1991).

A congenital or developmental form due to a neuromigration defect has been recognized and is known as the opercular syndrome (Graff-Radford et al., 1986; Becker et al., 1989) or congenital bilateral perisylvian syndrome (CBPS; Kuzniecky et al., 1993, 1997). In this instance, bilateral perisylvian polymicrogyria is seen on MRI scan and has been confirmed on post-mortem in some cases. It is of interest that the perisylvian syndrome has been described as a syndrome of childhood-onset epilepsy which, although sharing features of WDS, was reported without reference to that condition.

Clinical features of pseudobulbar palsy

Most of the descriptions of pseudobulbar palsy are in the adult literature and are due to acquired lesions. The fact that a child is evolving through neurological maturational phases and is similarly acquiring developmental skills means that expression of any impairment may vary or even be silent, depending on the particular stage of the child. Furthermore, developmental dyspraxias and maturational delays can present in superficially very similar ways.


Speech characteristics in adults with spastic dysarthria include a strained–strangled voice quality, hypernasality and consistent misarticulations with imprecise consonant production. Swallowing may also be abnormal and characterized by oral and pharyngeal stage difficulties. As a result, meal times may be prolonged and some patients are at risk of aspiration/penetration of food/liquid into the airway.


In the most severely affected cases, babies may present from birth with airway and sucking difficulties. For others, the problems with lip closure, tongue mobility and palatal function become apparent with weaning when they are unable to manipulate, transfer or chew food. Another large group present predominantly with speech delay. Persistent dribbling is very common. Many affected children can experience significant problems with eating, drinking and speaking throughout their lives, although, for some, the difficulties are reported to remit with time (examples of varying severity are given in the Appendix).

Clinical features of dyspraxia and maturational delay affecting oromotor function

Dyspraxia is difficulty with complex movement and motor planning that is not secondary to paralysis, weakness, incoordination, sensory loss or comprehension impairments. There are presumed deficits in the cortical motor association areas (and structural lesions have been demonstrated in adults). Oromotor dyspraxia causes inconsistent disturbances in articulation, phonation, respiration and resonance, leading to speech that is markedly dysfluent and unintelligible.

Developmental dyspraxia relates to difficulty with motor performance that is substantially below that expected given the child's developmental stage and cognitive ability. In speech dyspraxia, the pattern of speech is deviant and may be characterized by bizarre omissions and inconsistent substitutions, and specific difficulty with word synthesis. These children may also have a history of early feeding difficulty, but generally speech is the main bulbar modality that is affected. The issue of whether these patients are separable from those with WDS has not been examined and is not the purpose of this study.

In some children, there appears to be delay in maturation of the motor learning of articulation. There will be a normal period of babbling, but they will be slow to say their first words, will have fewer words in their spoken vocabulary and these words will be pronounced in an immature way.


The purpose of this study was to clarify the diagnosis of WDS and to record associated features and complications.

Subjects and methods


The first author (M.C.) recruited 47 children into the study from the caseloads of neurologists and therapists practising at a tertiary centre known to have a special interest in this condition. The entry criterion was that the major motor impairment was an upper motor neurone bulbar palsy, but excluding pure speech impairment without neurological signs and other evidence of bulbar dysfunction (i.e. excluding pure speech dyspraxia). The main differential diagnoses encountered were maturational delay in association with learning difficulties, oromotor dyspraxia and syndromic diagnoses such as velocardiofacial syndrome.

For each subject enrolled into the study group, a detailed case-note analysis was completed, including sections on possible aetiological factors, clinical features and relevant investigations. From the retrospective data available, it was not possible to interpret consistently the severity or anatomical extent of the pseudobulbar dysfunction, although all children had neurological signs involving at least one site (e.g. tongue, lips, jaw or soft palate). Similarly, there were insufficient serial examinations recorded to evaluate any tendency to improvement, although there were cases in which this certainly occurred from the clinical description available.

Statistical analysis

Descriptive statistics were used to examine each study variable individually. If an item of information was not available or not known for a particular case, the case was removed from the analysis of that item; the denominator in all items is thus the number of valid data points. Cluster analysis helped to examine subgroupings using a dendogram. Correlations between key variables were also analysed, and significance assessed using a Pearson two-tailed test.


The study group comprised 47 children (35 boys, 12 girls). The results for individual variables generally are expressed with numerator and denominator to indicate completeness of the data set. The ages of the children when first seen ranged from neonates to 13.8 years, with a mean age of 5.5 years. This was not a longitudinal study; however, many of the children have been followed for several years. The average age at last review was 7.3 years, and included three children originally seen as neonates, now aged 2, 7 and 8 years, and a 19-year-old who had been followed for 14 years. The majority were referred initially to neurology out-patients (26/45, 58%), others to the communication aids clinic (8/45, 18%) and a few to the neurodevelopmental clinic (4/45, 9%). The remainder had been admitted as in-patients (4/45, 9%).

Twenty-four (24/46, 52%) of the study group had a significant presentation to medical services in the first year of life, and in most instances this was due to feeding problems. However, the mean age at diagnosis was 6 years. Practically all the children had a statement of educational need (the UK method of designating children who need extra help in school).

In Table 1 we describe the main clinical features of our population and contrast these with the series obtained by Worster-Drought in 1956 and 1974 (Worster-Drought, 1956, 1974), and with the study of CBPS (Kuzniecky et al., 1993).

View this table:
Table 1

Main clinical features of the study population and comparison with the series obtained by Worster-Drought in 1956 and 1974 and with the study of CBPS ( Kuzniecky et al., 1993 )

Current studyWDS (1956)WDS (1974)CBPS (1993)
No. of patients478220031
% Male35/47 (74%)11/31 (35%)
Familial cases6/46 (13%)17/82 (21%)12/200 (6%)2/31 (6.5%)
Family history of neurological disease5/46 (11%)1/31 (3%)
Normal pregnancy36/45 (80%)25/31 (81%)
Normal delivery41/45 (91%)31/31
Neonatal complications12/45 (27%)3/31 (10%)
Epilepsy13/47 (28%)7/200 (3.5%)27/31 (87%)
Normal intellect9/47 (19%)60/82 (73%)122/200 (61%)5/31 (15%)
Moderate learning difficulties23/47 (49%)12/82 (15%)64/200 (32%)23/31 (75%)
Severe learning difficulties15/47 (32%)10/82 (12%)14/200 (7%)3/31 (10%)
Congenital defects31/47 (66%)9/82 (11%)48/200 (24%)10/31 (30%)
Palatal, cleft27/200
arythrogryposis multiplex14/31
Club foot26/31
Fused teeth2
Abnormal EEG8/23 (35%)27/31 (85%)
Receptive language commensurate with cognition45/46 (98%)31/31
Expressive language below receptive43/46 (93%)31/31
No speech19/46 (41%)2/31 (6.5%)
Microcephaly7/39 (18%)5/31 (16%)
Pyramidal signs41/45 (91%)22/31 (71%)
Abnormal eye movements6/46 (13%)
Abnormal neuro-imaging12/37 (32%)31/31
MRI bilateral perisylvian polymicrogyria5/37 (14%)31/31

Possible aetiological factors

Four (4/45, 9%) of the mothers had small vaginal bleeds in early pregnancy and 3/45 (7%) had some hypertension towards the end. However, 41/45 (91%) went on to have a normal vaginal delivery.

Three (3/47, 6%) babies were premature (one at 30 weeks, two at 35 weeks gestation). Twelve (12/45, 27%) of the babies needed neonatal care; for 10 of these babies, this was directly attributable to the pseudobulbar palsy (causing feeding and airway problems), and in no subject was perinatal asphyxia recorded. Five (5/45, 11%) babies were less than the third centile for gestational age. One (1/21) had an abnormal chromosome (fragile site on long arm of chromosome 2).

Six (6/46, 13%) children had a family history of WDS, 4/46 (9%) of undefined speech problems and 5/46 (11%) of neurological problems (two with epilepsy, two with sensorineural deafness, one with learning difficulties).

Feeding and its complications

Twenty-three (23/46, 50%) of the study group presented with feeding problems and 28/47 (60%) were recorded as having feeding difficulty from birth on direct questioning. Twelve (12/47, 26%) required tube feeding (either nasogastric or gastrostomy) at some stage. At the latest assessment, 6/47 (13%) children still relied entirely on tube feeds and only 9/47 (19%) took a completely normal diet including a full range of textures. The majority (25/47, 53%) still required a soft or carefully prepared diet. Only three children were documented as never having had feeding problems.

Gastro-oesophageal reflux was recognized frequently, being documented in 18/44 (41%). Ten (10/44) were managed medically and 8/44 (18%) had surgical correction. Poor nutrition was common, with 14/38 (37%) having a history of failure to thrive (weight less than the third centile for age, or of crossing the centiles), and 6/39 (15%) weighed less than the third centile for age at their most recent assessment. This was despite the fact that just 11% had birth weights at or below the third centile for gestation.


In almost all cases (45/46, 98%), receptive language correlated with non-verbal understanding and there was a characteristic and significant gap between receptive and expressive language in 43/46 (93%). Expressive speech problems ranged from particular difficulty with lingual and labial sounds with hypernasality, to severe cases where phonation itself was affected due to impaired laryngeal muscle control. Nineteen (19/46, 41%) of the children had no speech just simple vocalizations, 18/46 (39%) were intelligible only to familiar listeners and only 8/46 (17%) were intelligible to strangers.

Thirty-six (36/46, 78%) of the study group used augmentative communication methods to supplement or replace speech. Thirty-one (31/43, 72%) used manual signing, 18/43 (42%) a symbol-based system and 8/42 (19%) a voice output device.


Seven (7/45, 16%) children presented with neonatal respiratory problems (apnoea, respiratory obstruction or aspiration) directly attributable to the pseudobulbar palsy and requiring admission to a neonatal unit. Two children went on to have tracheostomies and gastrostomies to protect their airways. Sixteen (16/44, 36%) children of varying ages gave a history suggestive of aspiration.

Secretory otitis media occurred in 22/36 (61%) children, with 18/36 (50%) receiving grommets.

Dribbling and dental

Dribbling was commented on specifically in 37/43 (86%) case notes and was often accompanied by an open-mouthed posture and flat impassive facies. In nine cases, the child had tried medical treatment and five had submandibular duct transposition.

Nine (9/43, 20%) study children were recorded with dental problems, including two children with fused teeth.

General development and cognitive level

Thirty-three (33/45, 73%) study children had delay in developmental milestones (other than language). They sat on average at 11 months (31 data points), walked at 2.1 years (36 data points) and were toilet-trained by day at ~3.5 years (only nine data points).

Nine (9/47, 19%) were of normal intelligence, 23/47 (49%) had moderate learning difficulties (IQ 50–70 range) and 15/47 (32%) had severe learning difficulties (IQ < 50).


Three (3/47, 6%) had possible neonatal seizures (one secondary to apnoeas, one due to hypocalcaemia, and one had an unexplained eye-rolling episode at 12 days). Four (4/47, 9%) had febrile fits.

Thirteen (13/47, 28%) of the study group had epilepsy. Seven had weekly or daily seizures and five were resistant to multiple anticonvulsant combinations. Five had had status epilepticus.

Psychiatric disorders

Nineteen (19/46, 41%) were recorded with psychiatric or emotional problems, with four of these children having more than one type of problem. Attention deficit hyperactivity disorder occurred in eight children (five of whom were on treatment with methylphenidate). Four children were within the autistic spectrum. Nine were noted to have a conduct disorder and four were described as having a sleep disorder.

Clinical findings

Oromotor dysfunction was documented in all children, with abnormal tongue movements being most common and present in 41/42 (98%), followed by abnormal lip movements in 37/38 (97%), a brisk jaw jerk in 35/39 (90%) and drooling in 32/40 (85%). Some details were recorded in comparatively few case-notes, for example the soft palate appeared abnormal in 20/23 (87%) and jaw movements were deviant in 9/12 (75%).

Forty-one (41/45, 91%) had some limb pyramidal features. In most cases (40/45, 89%), this was brisk reflexes, with 20/40 (50%) having increased tone, 15/27 having extensor plantars and 12/39 (31%) having clonus. Seven (7/39, 18%) were microcephalic (head circumference less than the third centile for age).

Thirty-one (31/47, 66%) had congenital defects (see Table 1), particularly affecting mouth structures (including four with jaw contractures and two with cleft palates) and limbs (predominantly contractures with one child having multiple congenital joint contractures, i.e. arythrogryposis multiplex).

Six (6/46, 13%) children were thought to have abnormal eye movements (not including those with simple squints). No other additional cranial nerve abnormality was noted.


Thirty-seven children had neuroimaging, of whom five had only a CT scan, eight had a CT and MRI and the remaining 24 had an MRI alone. The majority of MRI scans used the routine sequence of the 1.5 Tesla MR Unit at Great Ormond Street Hospital and were reviewed by a consultant paediatric neuroradiologist.

Two abnormalities were detected on CT scan alone (one showed mild right hemi-atrophy and one had an area of low density lateral to the left lateral ventricle in the temporo-parietal region)

Ten of the MRI scans were abnormal. Five had bilateral perisylvian polymicrogyria. The other five comprised: (i) polymicrogyria, subependymal grey matter heterotopia and partial agenesis of the corpus callosum; (ii) mild cerebral atrophy; (iii) moderate cerebellar atrophy with poor grey–white differentiation and an abnormal high signal from the white matter of the external capsule on T2; (iv) a small pons and asymmetric cerebellar hemispheres (left smaller); and (v) two small foci of abnormal signal in the left frontal white matter and adjacent to the left head of the caudate nucleus.

Nineteen of the children had a documented videofluoroscopy, and 18 of these studies were abnormal, usually involving multiple phases of the swallow, particularly the oral (14) and pharyngeal (11) phases. In six studies, there was clear aspiration.


Main findings

WDS was described originally as an isolated congenital pseudobulbar paresis. The most striking finding of this study is the high frequency and wide range of additional impairments, which tend to dominate management: 91% had some pyramidal features; 81% had learning difficulties; 66% had congenital defects; 41% had neuropsychiatric problems; and 28% had epilepsy

The pseudobulbar paresis in itself had severe direct consequences, with 60% having feeding difficulties from birth and 81% still not managing a completely normal diet at their most recent assessment (age range 1–19 years, with 45/47 being older than 2 years). Similarly, 93% had a significant difference between their receptive and expressive language, with 78% using augmentative communication methods when last reviewed. Secondary complications were also very significant.

Comparison with other studies

Our study group contrasted with that originally described by Worster-Drought in that he reported a largely isolated pseudobulbar palsy (although his later study acknowledged additional impairments). Instead, our study children have complex disabilities. This discrepancy supports the contention that there is a selective emphasis on bulbar motor impairment in this group despite the more widespread additional impairments. The focus on bulbar difficulties may also be the reason that these children have not been recognized within the cerebral palsies as they present initially for feeding and language management, with other impairments only becoming apparent with time.

In his original paper, Worster-Drought reviewed 82 cases, with seven families contributing 17 cases (21%). Nine children (11%) had congenital defects and 22 (27%) learning difficulties. In 1974, Worster-Drought described a further series of 200 cases in which he recognized an increased incidence of learning difficulties (78/200, i.e. 39%, having an IQ < 80), and an association between learning difficulties and greater severity in the pseudobulbar palsy. There was also a higher incidence of congenital defects than previously recognized (48/200, 24%) and, in this series, Worster-Drought acknowledged that traditional speech therapy appeared ineffective in the more severely affected children.

Worster-Drought's series of cases were derived from a caseload of children seen in a special school that did not take children with major learning problems. This has almost certainly resulted in some bias in the sample towards children with milder cognitive impairment and fewer additional impairments than otherwise. He was also able to identify the condition on the basis of a phenotype without the help of modern imaging.

CBPS appears to fall within the definition of WDS. The CBPS study (Kuzniecky et al., 1993) was not population based, but was drawn from a sample of patients undergoing clinically indicated neuroimaging, biasing the sample towards patients where imaging may affect management (e.g. severe epilepsy under consideration for surgery or arythrogryposis multiplex). There was no opportunity to ascertain whether the imaging changes were essential to the clinical phenotype described. Despite the above reservations, our study findings are strikingly similar to those of the CBPS study, with learning impairments, pyramidal signs and epilepsy featuring strongly, and neither study finding obvious risk factors, apart from a positive family history in a significant minority. The `epilepsy rate' was lower in our study, which may reflect differences in sample bias or younger age in our study group (the CBPS study suggested that the seizures often started in the 6–12 year age range).

Aetiology and pathogenesis

In this series, there were no obvious causes in the pregnancy or birth, and possible risk factors in a minority. This is unlike some forms of cerebral palsy, e.g. spastic diplegia (which is highly correlated to pre-term birth) and athetosis (which characteristically follows full term birth asphyxia), as described in the Gothenberg series (Hagberg and Hagberg, 1984). In the Gothenberg series (which incidentally is one of those in which WDS was not recorded), multiple risk factors were found, albeit that they were asked for systematically.

In the original description, it was postulated that WDS was due to a `developmental defect of the corticobulbar neurones'. The clinical findings of abnormal eye movements in a minority of our study children might suggest lesions in the anterior internal capsule (but this is not strong evidence and no such capsular lesions have been reported). However, advances in imaging mean that we now know that some of these children have clear neuromigrational changes in the perisylvian regions. Moreover, the common association of a tetraplegia, cognitive problems, epilepsy and congenital abnormalities suggests that the abnormality is not confined to a circumscribed lesion in the nervous system, but involves more widespread effects and probably often occurs in early fetal development. Evidence for a middle cerebral ischaemic lesion involving the perisylvian region during the phase of neuronal migration between 12 and 16 weeks gestation has been proposed (Meencke and Veith, 1997).

Teasing out these factors is difficult, for instance in one study child there were MRI changes of bilateral perisylvian polymicrogyria, whereas his father had WDS on clinical grounds, but a normal brain scan. The excess of boys in our study group may suggest a particular vulnerability (although this sex ratio was reversed in the CBPS study). In Patton's study, one of the cases of familial WDS appears to have developed normally until an acute illness in the second year left him with a pseudobulbar palsy (Patton et al., 1986), in contrast to Worster-Drought's usual exclusion of `acquired' cases. The cause of this illness is unclear, although the child is reported to have regressed in many areas of development and to have seizures, so that one might propose an epileptic regression in association with non-convulsive status. One of our study children also had documented normal development until his second year, when he regressed in association with seizures and non-convulsive status, which has been reported in the perisylvian syndrome. These observations suggest that in some instances, the condition may be due to a gene defect with variable penetrance that predisposes an individual to pseudobulbar palsy, the expression of which might depend on pre- and post-natal factors. Timing of the effect of the gene defect on opercular development may account for similar phenotypes in the same family but with different scan appearances.

As with other forms of cerebral palsy, the clinical phenotype of WDS may result from a variety of early CNS insults, and the original theoretical pathological separation by Worster-Drought is of doubtful relevance. The presence of cognitive, behavioural and seizure impairments strongly supports the cerebral cortical and presumably perisylvian localization. These three impairments have been regarded as clustering in association with congenital cortical grey matter damage (Neville, 1993).

Evidence for subgroupings and prognostic measures

Many of the variables were strongly interrelated as they measured different aspects of a given difficulty experienced by a child. This was examined using simple bivariate correlations (significance assessed with a Pearson two-tailed test), and it confirmed associations such as: (i) children with severe feeding problems tend to present as neonates or in the first year and they experienced complications such as failure to thrive, gastro-oesophageal reflux, aspiration, current weight less than the third centile and glue ear; and (ii) poor expressive language was associated with use of alternative communication methods, cognitive problems and ongoing feeding difficulties.

These types of correlations confirmed the need to anticipate complications and reflected practical aspects of expressive speech problems, but did not clarify any underlying associations amongst these children. Therefore, hierarchical cluster analysis was used to reveal meaningful groupings within the children.

  • (i) A small number of children (5/47, 11%) presented with severe feeding problems, usually as neonates, or certainly within the first year. Most needed tube feeds (4/5) initially and many (3/5) continued to be fed in this way. They tended to experience failure to thrive, gastro-oesophageal reflux (usually requiring surgery) and aspiration and went on to have major expressive language problems when older, needing high levels of augmentative communication support. This set appeared cognitively more intact than the overall group, albeit few in number (2/5 normal intellect, 2/5 moderate learning difficulties and 1/5 severe learning difficulties).

    Presumably this group reflected those who, in Worster-Drought's original description, had a complete or severe pseudobulbar palsy. The retrospective case-note review does not give sufficient detail on site or severity of the pseudobulbar palsy to clarify this.

  • (ii) Another relatively well-defined group (12/47, 26%) tended to present later with developmental or language problems and had only mild or moderate feeding problems. All of these children tended to have moderate to severe expressive speech problems (from intelligible to familiars only, to no speech) and required communication aids. There was also a strong tendency towards learning difficulties (11/12, 92%; with 5/12 severe learning difficulties and 6/12 moderate learning difficulties).

  • (iii) The remaining set (30/47, 63%) often presented with moderate feeding problems that showed a tendency to improve such that they could cope with a `careful' diet (e.g. avoiding meat and crisps). However, with time, expressive speech difficulty became the predominant problem, with most needing some augmentation. Cognitive levels reflected those of the overall study group.

It was also thought useful to see if any symptoms were reliable predictors of the more severe outcomes. Multiple regression analysis was used to look at how far current feeding, expressive language, cognition and epilepsy could be predicted from early symptoms. In the case of current feeding, three factors (aspiration, gastro-oesophageal reflux and presenting in first year of life) consistently accounted for ~55% of the variance. For expressive language, the best predictors (32% explanation) were gastro-oesophageal reflux and cognitive level, and for cognitive level the predictors were bilateral perisylvian polymicrogyria, developmental presentation and expressive language (28% explanation). From the symptom data available, it was not possible to predict epilepsy. Examples of cases of differing severity are given in the Appendix.

Clinical implications

The practical importance of this condition, which we feel is easy to recognize, is in the complexity of the management due to the multiple small impairments (previously explored with the parent support group; Neville, 1997), e.g. the use of alternative communication systems is hampered by attention, cognitive and manual coordination impairments.

One of the striking findings from this study is that the majority of children presented to medical attention by 12 months, but the diagnosis of WDS was not confirmed until a mean age of 6 years. The study also highlights the complexity and severity of disabilities affecting these children. Even if the pseudobulbar palsy was considered in isolation for its immediate mechanical implications, an outcome where 40% of cases have no speech and 80% need a modified diet would be considered poor on any rating, let alone the fact that 40% have gastro-oesophageal reflux and 36% have aspiration. Add to this the widespread associated impairments found in this study group (such as 80% learning difficulties, 55% psychiatric disorders, 28% epilepsy, and poor motor and organizational skills in the majority) and it is easy to see why they are challenging children to work with.

The multiple difficulties compound each other, and often test the professional team. No area or skill can be tackled in isolation as they are all interdependent and need an integrated approach. For example, the response to conventional speech and language therapy is controversial and not proven for these children. It is vital to provide them with early alternative communication methods to support their cognitive and psychosocial development. However, such children will have to overcome enormous obstacles even to use alternative methods, as they are often additionally disadvantaged with learning difficulties, poor attention span and poor motor organizational skills, making suitable educational support very complex. Furthermore, they are ambulant so that any augmentative communication equipment needs to be robust and highly portable.

In addition, there are hidden consequences of these complex disabilities. It is recognized from other studies (Boyle, 1991; Reilly et al., 1996) that poor nutrition affects growth and general health, lowers the immune response, reduces attention span and learning ability, decreases motivation, increases irritability and can adversely affect early parent–child interaction (as meal times become stressful and unpleasant occasions for the family and the child can never be left alone with a snack). Many of these factors will be having their strongest effect in the first few years of life when feeding is at its worst. However, this is probably the most vulnerable period in development and it may be difficult to compensate for negative experiences in areas that would normally facilitate development, such as attention, motivation, mood and social interaction.

Similarly, in areas of communication, it seems that many of the children are late in developing a means of expression commensurate with their level of understanding. Theoretically, it seems likely that this lack of early functional communication (particularly outside the immediate family) would have major effects on motivation for social interaction, self-esteem and levels of frustration. Although higher rates of behavioural and emotional disorders in children with neurological impairment are a well-established finding (Goodman, 1993) (also seen in our study) and emotional lability is a common association in adults with pseudobulbar palsy, an additional factor for our children could be their difficulties with early communication and interaction.

It is vital that these children are managed by a multidisciplinary team, with a wide range of expertise including medical, psychological, psychiatric, communication and dysphagia interests, which can anticipate and address these diverse needs appropriately.

It is particularly important not to miss treatable complications such as gastro-oesophageal reflux, aspiration and malnutrition which all affect functioning. Similarly, many of the psychiatric disorders will respond to medication, such as methyl phenidate for attention deficit hyperactivity disorder, and behavioural programmes, e.g. for austistic spectrum disorders. Optimizing seizure control is also essential to minimize the effects of both fits and anticonvulsants on information processing and behaviour, and the clinician must be alert to the potential for non-convulsive status which can cause plateauing and regression.

Methodological issues

This was not a population-based study and it is likely that there was sample bias towards severe cases by the referral characteristics and known expertise of a tertiary centre with a special interest in this group of conditions. The sample size was small and lacked suitable controls so that it is not possible to draw aetiological conclusions or clearly reach definitive diagnostic criteria. The study also relied on retrospective data that could not be tested formally for reliability or validity, although they were in the main collected by our group. Finally, there are no clear guidelines for diagnosing pseudobulbar palsy in children where cooperation with examination cannot be assumed and the distinction from oromotor dyspraxia is difficult.

Despite the above comments, it is essential that there is some attempt to review this unusual and important group of children.


WDS falls clearly within the rubric of the cerebral palsies as a syndrome that includes motor impairment and arises from static damage to the brain in early life. The presence of primary grey matter impairments (cognitive, behavioural and epilepsy) and its motor distribution with a caudo-cephalic sequence of increasing severity of pyramidal involvement puts this clearly into the category of a mild spastic tetraplegia. This relationship is clear providing the strict definition of spastic tetraplegia from the Gothenburg group is used, i.e. pyramidal involvement, maximum in bulbar structures and upper limbs, with major cognitive impairment, and a very high rate of epilepsy with bilateral widespread cerebral cortical damage. This should not be confused with four limb and sometimes bulbar involvement in the severe end of the spastic `diplegia' sequence of prematurity due to periventricular leucomalacia.

Epilepsy as a feature of this syndrome is important as part of the cortical grey matter impairment complex but it also shows the potential for non-convulsive status with cognitive, behavioural and bulbar plateauing and deterioration, a factor which should be investigated properly.

We would propose the following as core elements for WDS (see Table 2).

View this table:
Table 2

Suggested criteria for diagnosis of WDS*

Supportive featuresFrequency (%)Discriminatory power
*The absolute criterion for diagnosis of WDS is clinical signs of a pseudobulbar paresis which is both congenital and the major motor impairment. A difficulty with constructing criteria for WDS is that many features may be revealed with time (including the pseudobulbar paresis if you allow cases that become manifest at the time of an epileptic regression). The scheme above sets out frequency and discriminatory power for various characteristics based on our study.
History of delay in gross motor function beyond that expected by cognitive level90High
Mild pyramidal features on examination90High
Expressive language level significantly below that of receptive language90High
Drooling inappropriate for developmental stage85High
History of feeding problems (may be subtle)80High
Learning impairment80Low
Glue ear60Low
Recurrent respiratory infections/aspiration40Low
Gastro-oesophageal reflux40Low
Neuropsychiatric problems40Low
Bilateral perisylvian polymicrogyria15High

  • (i) Congenital suprabulbar paresis causing feeding problems, speech difficulty and drooling with secondary effects of inhalation, middle ear disease (due to poor Eustachian function) and gastro-oesophageal reflux (due to lack of regular normal swallowing causing disordered oesophageal motility—the alternative postural explanation for gastro-oesophageal reflux because of the profound motor delay that characterizes spastic tetraplegia is not supported in this case because the gross motor delay in the WDS is mild).

  • (ii) Mild/moderate motor delay with evidence of pyramidal disease which, in many, then causes a pathological delay of several months between walking while holding on (`cruising') and stable independent walking.

  • (iii) A significant excess of cognitive and behavioural impairments and epilepsy which leads us to conclude that the majority of the causes are due to cortical grey matter damage presumably in the perisylvian region (Neville, 1993). These impairments pose many of the difficulties in management of these children and adults.

The history of this group of disorders is fascinating. Despite the clear description of the condition more than 40 years ago, it has been startlingly absent from epidemiological studies of the cerebral palsies, e.g. the Gothenburg study (Hagberg and Hagberg, 1984), South East Thames (Evans et al., 1985), Danish (Glenting, 1976) and the UK (Pharoah et al., 1998), presumably because of lack of clinical awareness of the condition and the lack of major gross motor impairments. In our experience, the condition is not uncommon. The description of the bilateral perisylvian syndrome as a completely separate condition dominated by intractable epilepsy mirrors the over-selective approach many WDS children receive through presenting initially for feeding and language management, and highlights the importance of a holistic developmental and neurological approach to complex children. The total overlap of the phenotypes between the bilateral perisylvian syndrome and our version of WDS leads us to propose that they are the same condition and that the disturbance in some cases may be at a functional rather than anatomical level, such that obvious perisylvian polymicrogyria on MRI need not be present. The occurrence of patients within the same family with and without such imaging abnormality supports this view. It is likely that timing of the perisylvian developmental damage may influence the gross anatomy of the cortex but leave the functional impairments very similar. From the reports available, one might conclude that the obvious perisylvian polymicrogyria is more likely to be associated with epilepsy, but the ascertainment bias of epilepsy units is a much more likely explanation.

Future studies

Further study of children with WDS would offer an important chance to understand more about bulbar function generally in infancy and childhood, and may thereby help to clarify difficulties experienced by large numbers of children with other types of cerebral palsy. It may also shed light on aetiological mechanisms, prognostic features and functional outcome in this form of cerebral palsy, and give a logical basis on which to plan treatment (e.g. with current knowledge, decisions about effective communication support for these children, whether traditional speech therapy or augmentative methods, are slow and largely by trial and error; understanding the site and severity of the pseudobulbar palsy and its potential for improvement would help to clarify this).

A prospective trial of early intervention in areas of feeding and communication would also be important, as would a review of adult outcome, with particular regard to functional adjustment and independence.

Following this retrospective review, we plan to undertake a prospective study of these children with standardized history and assessment to clarify many of these issues.

Appendix: Examples of differing severity


Case A presented from birth with recurrent apnoea and aspiration requiring ventilation. He needed a tracheostomy and epiglottoplexy. A has never tolerated oral feeds and has also had major problems with gastro-oesophageal reflux, managed with a gastrostomy and a Nissen's fundoplication. A does not vocalize, but uses augmentative communication systems including Makaton, picture symbols and an Alpha Talker. He has severe learning difficulties and a particularly notable delay in his motor development (walked at 3 years 10 months with frame, at 7 years independently).

On examination, A has constant and profuse drooling, reduced facial movements, limited jaw movement, poor lip closure, minimal tongue movement, an absent gag and swallow, and a brisk jaw jerk. He has little voluntary eye movement. Tone is generally increased with brisk reflexes, with patterning of hand movements severely limiting function.

MRI, EEG and routine karyotype were normal.


Case B was a slow feeder as an infant (taking 2 h per feed) and had problems with weaning and lumpy food. Even at 8 years he was only able to manage small soft lumps (e.g. chopped fish finger, banana), and dribbling was still constant. He had recurrent chest infections and all growth parameters (including head circumference) are below the second centile.

Language development was delayed and deviant, with B acquiring babble at 2 years, and first words at 6 years. His speech (at 8 years) is intelligible only to familiar people, and he relies on Makaton. He has severe learning difficulties, and delayed motor development, sitting at 19 months and walking at 3 years. B also has major behavioural problems including aggressive outbursts (necessitating residential care) and attention deficit (with trial of methyl phenidate). He has complex partial seizures (up to 15/day) which have been resistant to treatment.

On examination, B is constantly active with poor attention and poor cooperation. He has profuse dribbling, reduced oral movements and a brisk jaw jerk. He has generally increased tone with brisk reflexes and ankle clonus.

MRI and chromosomes were unremarkable, and EEG showed paroxysmal bursts.


Case C was always slow to feed as an infant, and had problems with chewing and clearing his mouth as a young child, occasionally inhaling food. By 6 years, he could manage a careful diet, avoiding meat. He dribbles to a varying degree, particularly when concentrating.

C has moderate learning difficulties, with comparable receptive language. However, expressive language is significantly worse, developing first words around the age of 4 years, and short sentences by 6 years. He is understood only by familiar people, and also uses signing and symbol books. C walked at 17 months and has always been seen as `clumsy', still tripping frequently at 6 years and not being able to do up buttons.

C has not had obvious seizures, although he has been investigated for eye-rolling episodes. He tends towards stubborn and oppositional behaviour. There is no family or perinatal history of note.

On examination, C is excitable with variable dribbling. He has slowness of facial and tongue movements, with a normal jaw jerk. Tendon reflexes are generally increased, with unsteadiness of hand function without obvious signs. MRI and EEG are normal.

Epileptic regression

Case D had normal early development, no problems with early suckling or weaning, sat at 7 months, walked at 13 months and had a few clear single words around the age of 1 year. At 15 months, he developed `episodic rigidity', followed by regression particularly in language and motor skills. He then had frank generalized tonic–clonic seizures and intermittent non-convulsive status that was largely refractory to anticonvulsants (including steroids).

D continued to have an intractable seizure disorder, with difficulty with motor organizational skills of all types, particularly oromotor, and has severe learning difficulties. At review at 12 years he needed food to be cut up as he could not chew, regularly coughed or choked during meals and often regurgitated. His speech comprised largely vowel and glottal sounds and he had difficulty using sign language because of poor motor organizational skills. D's non-verbal and receptive language understanding were the age equivalent of 3 years.

On examination, D was open mouthed and drooled when concentrating. He had poor lip closure and could protrude his tongue to form a tip, but not elevate, depress or move it laterally. Reflexes were elicited easily, with extensor plantars and difficulty with fine movement. MRI was normal and EEG was always abnormal with varying severity.


The parent support group has been a significant source of information and encouragement. This work was undertaken by Great Ormond Street Hospital for Children NHS Trust who received a proportion of its funding from the NHS Executive; the views expressed in this publication are those of the authors and not necessarily those of the NHS Executive.


  • Present address: Faculty of Health Sciences, La Trobe University, Melbourne, Australia


View Abstract