Aims The Tuberous Sclerosis 2000 Study is the first comprehensive longitudinal study of tuberous sclerosis (TS) and aims to identify factors that determine prognosis. Mode of presentation and findings at initial assessments are reported here.
Methods Children aged 0–16 years newly diagnosed with TS in the UK were evaluated.
Results 125 children with TS were studied. 114 (91%) met clinical criteria for a definite diagnosis and the remaining 11 (9%) had pathogenic TSC1 or TSC2 mutations. In families with a definite clinical diagnosis, the detection rate for pathogenic mutations was 89%. 21 cases (17%) were identified prenatally, usually with abnormalities found at routine antenatal ultrasound examination. 30 cases (24%) presented before developing seizures and in 10 of these without a definite diagnosis at onset of seizures, genetic testing could have confirmed TS. 77 cases (62%) presented with seizures. Median age at recruitment assessment was 2.7 years (range: 4 weeks–18 years). Dermatological features of TS were present in 81%. The detection rate of TS abnormalities was 20/107 (19%) for renal ultrasound including three cases with polycystic kidney disease, 51/88 (58%) for echocardiography, 29/35 (83%) for cranial CT and 95/104 (91%) for cranial MRI. 91% of cases had epilepsy and 65% had intellectual disability (IQ<70).
Conclusions Genetic testing can be valuable in confirming the diagnosis. Increasing numbers of cases present prenatally or in early infancy, before onset of seizures, raising important questions about whether these children should have EEG monitoring and concerning the criteria for starting anticonvulsant therapy.
Statistics from Altmetric.com
Tuberous sclerosis (TS) is an autosomal dominant disease characterised by tumour-like lesions in the skin, brain and other organs.1 The major diagnostic features are listed in box 1. Most children with TS have epilepsy and about half have intellectual disability. Autism spectrum disorder and attention deficit hyperactivity disorder are common. The condition is caused by mutation of the TSC1 or TSC2 gene. Rare patients with deletions of TSC2 that encompass the adjacent PKD1 gene present with severe early onset polycystic disease.2
What is already known on this topic
▶ Children with tuberous sclerosis (TS) who are too young to manifest characteristic skin lesions may not fulfil the clinical criteria for a definite diagnosis.
▶ Most children with TS present with seizures in infancy or early childhood.
▶ Seizures may contribute to cognitive impairment in children with TS and early treatment may improve cognitive outcome.
What this study adds
▶ Genetic testing should be more widely used for confirming the diagnosis, particularly in young children.
▶ Identification of abnormalities at routine antenatal ultrasound examination is now the commonest mode of presentation of TS after seizures.
▶ Diagnosis of TS before epilepsy onset is common, raising difficult questions about monitoring for seizures and indications for starting anticonvulsant therapy.
TS shows extraordinary variation in severity and manifestations, depending on which systems are involved and at what age. This causes difficulties with diagnosis and management. Diagnosis is usually based on clinical and radiological findings.1 3 One major feature raises the possibility of TS and two major features are needed for a definite diagnosis (box 1). Genetic testing identifies mutations in most children with TS but is not widely used in the UK for confirmation of the diagnosis.
Major features of tuberous sclerosis (modified from Roach et al3)
▶ Two major features: definite diagnosis
▶ One major feature: possible diagnosis
▶ Facial angiofibromas or forehead plaque
▶ Ungual fibroma, non-traumatic
▶ Hypomelanotic macules, three or more
▶ Shagreen patch
▶ Multiple retinal nodular hamartomas
▶ Cortical tuber*
▶ Subependymal nodule
▶ Subependymal giant cell astrocytoma
▶ Cardiac rhabdomyoma, single or multiple
▶ Renal angiomyolipoma or pulmonary lymphangiomyomatosis†
*When cerebral cortical dysplasia and cerebral white matter migration tracts occur together, they should be counted as one rather than two features of tuberous sclerosis.
†When both lymphangiomyomatosis and renal angiomyolipomas are present, other features of tuberous sclerosis should be present before a definite diagnosis is assigned.
The literature on TS is dominated by clinic based studies which are inherently biased towards more severe cases. Population based studies avoid this bias but are few in number. The Tuberous Sclerosis 2000 Study is the first formal comprehensive longitudinal study to be carried out. Children aged 0–16 years were recruited as soon as possible after diagnosis so that good quality data on early seizure history and development could be collected in order to investigate the determinants of intellectual, cognitive and neuropsychiatric outcome. Here we report on the mode of presentation and findings at the initial assessments and discuss their implications for diagnostic evaluation and management.
The study was designed and monitored by a National Coordinating Committee of health professionals with a specialist interest in TS and patient representatives from the Tuberous Sclerosis Association. Children aged 0–16 years resident in the UK and diagnosed with definite or possible TS between 1 January 2001 and 31 December 2005 were ascertained by mailing all paediatricians, paediatric neurologists and clinical geneticists in the UK at the start of the recruitment period and annually thereafter. Cases were also ascertained through the Tuberous Sclerosis Association. Current diagnostic criteria were used3 and cases with a possible diagnosis of TS were included because young children with TS do not always meet the criteria for a definite diagnosis when they first present.
Children were reviewed by clinicians with experience of TS in a network of clinics covering the UK. At the initial recruitment assessment, a full medical history was obtained and a physical examination carried out including inspection of the skin with a Wood's lamp. A seizure history was obtained from the parents and cross-checked against the medical records. For a few cases unable to attend a specialist clinic, information was obtained from their clinician, the medical records and by telephone interview with the parents. Clinicians responsible for the routine care of the patients were provided with a list of recommended evaluations for the study (box 2) based on current guidelines.1 4 5 Cranial CT and MRI scans were independently reviewed by a neuroradiologist with particular experience of TS (JNPH). A team of clinical psychologists seeing families at home or in the clinic carried out age-appropriate intellectual, cognitive, behavioural and neurodevelopmental assessments. The results of the Mullen Scales of Early Learning6 and the Vineland Adaptive Behaviour Scale7 are reported here and were used to estimate intellectual ability and the presence of learning disability (estimated IQ less than 70). For assessment of learning disability precedence was given to the Mullen scales, and the Vineland score was only used when the Mullen was not available.
Recommended evaluations for the study
▶ Echocardiography in children under 5 years of age as part of the diagnostic work-up
▶ ECG as a baseline investigation to exclude cardiac conduction defects and arrhythmias
▶ Renal ultrasound scan at diagnosis for detection of polycystic kidney disease associated with contiguous gene deletions of the TSC2 and PKD1 genes
▶ Follow-up renal ultrasound scans at age 5, 8, 12 and 16 years as appropriate
▶ EEG for the evaluation of seizures
▶ Cranial MRI scan at age 2 years or older if not already done as part of the diagnostic work-up
Genotyping was carried out by the two diagnostic laboratories providing TS mutation testing in the UK (East Anglian Regional Genetics Laboratory, Cambridge and Institute of Medical Genetics, Cardiff). All exons and flanking intronic sequences of the TSC1 and TSC2 genes were screened for point mutations. Samples were tested for whole exon deletions (including TSC2/PKD1) by multiplex ligation-dependent probe amplification (MLPA; MRC-Holland, Amsterdam, The Netherlands).
We recruited 136 children thought to have a definite or possible diagnosis of TS. After critical review of the findings, three cases fell short of the criteria for possible TS and were excluded. Two families withdrew from the study. In six cases with possible TS at recruitment, the diagnosis remained unconfirmed and in four of these cases genetic testing including MLPA did not identify a pathogenic mutation. This report is therefore based on 125 children (62 male, 63 female) with a definite diagnosis of TS, including six pairs of siblings. One hundred and fourteen (91%) met clinical criteria for a definite diagnosis and in the remaining 11 (9%) genetic testing identified a pathogenic mutation which confirmed the diagnosis. In the 93 families with a definite clinical diagnosis who had genetic testing, the detection rate for pathogenic mutations was 89%.
Figure 1 shows the age distribution at presentation. For cases presenting postnatally, the median age at presentation was 7 months with no significant difference between males and females. Table 1 lists the modes of presentation and their frequency. Twenty one cases (17%) presented prenatally, usually with abnormalities identified at routine antenatal ultrasound examination, the commonest finding being rhabdomyomas. Including cases presenting at or soon after birth with a cardiac murmur or tachycardia, there were a total of 22 cases (18%) where the first indication of TS was the finding of single or multiple rhabdomyomas. A total of 30 cases (24%) presented before the onset of seizures. Of these, 23 have had genetic testing which identified a mutation, including 10 cases who did not have a confirmed clinical diagnosis when they had their first seizure.
The majority of cases (62%) presented with seizures in infancy or early childhood. The median age at presentation with seizures was 6 months (range: birth to 10 years) and the median time from presentation to diagnosis was 2 months (range: 0–14.9 years). Median age at recruitment assessment was 2.7 years (range: 4 weeks–18 years). Prior to recruitment or since entry into the study, 114 cases (91%) have developed epilepsy, 57 (50%) of whom have had infantile spasms.
Table 2 summarises the findings at examination and the results of investigations. The Vineland Adaptive Behaviour Scale and Mullen Scales of Early Learning were completed in 116 (93%) and 83 (66%) of the children, respectively. The results are given in table 3. There was good agreement between the two measures. Overall, 123 of the children had at least one of these assessments and 80 (65%) had some degree of intellectual disability.
This report is based on 125 children with a definite diagnosis of TS. We have not included six children who met the diagnostic criteria for possible TS but in whom further investigation was normal including genetic testing in four cases. These children are unlikely to have TS but their exclusion may have introduced bias.
In families with a definite diagnosis of TS on clinical criteria alone, genetic testing identified mutations in 89%, which compares favourably with other studies.8,–,10 Since mutations can be indentified in the majority of cases of TS, genetic testing can make a valuable contribution to the diagnostic evaluation,11 particularly in younger children who are less likely to fulfil the clinical criteria for a definite diagnosis. Clinicians in this study seldom used genetic testing as part of the diagnostic evaluation and this investigation was usually carried out later to address genetic issues in the family. In 11 cases who did not meet the clinical diagnostic criteria for TS, confirmation of the diagnosis for inclusion in this report came from genetic testing, which illustrates the value of this investigation.
The commonest mode of presentation in this study was with seizures in infancy or early childhood. Most cases were promptly referred to a hospital specialist and the median time taken to establish the diagnosis was 2 months. However, in a minority of cases it took much longer, as reported in previous studies.12 13 Clearly it is important to consider TS in young children presenting with seizures and to carry out a full diagnostic work-up. When this provides evidence of TS but does not meet the criteria for a definite diagnosis, genetic testing is indicated and can make a valuable contribution to confirming the diagnosis.
The number of cases presenting with abnormalities identified at routine antenatal ultrasound examination has increased in recent years and accounts for a substantial proportion of cases.13 14 In the current study it was the commonest mode of presentation after seizures. Cardiac rhabdomyomas were the most frequent finding. Including cases presenting at or soon after birth with cardiac signs or symptoms, there were a total of 22 cases (18%) where the first indication of TS was the finding of one or more rhabdomyomas. Multiple rhabdomyomas are highly likely to be associated with TS15 16 and the risk with single lesions, while lower, is still significant. In this situation it is important to consider TS and carry out a full diagnostic work-up, if appropriate delaying this until after birth for cases identified by antenatal screening. Here again, genetic testing has a valuable role in confirming the diagnosis.
Overall, 91% of cases had epilepsy with infantile spasms being common. Estimates of the frequency of seizures in TS vary widely, from 96% in one clinic based study17 to 79% in a population based survey.18
On physical examination, the commonest finding was hypopigmented macules, found in 78% of cases. Current diagnostic criteria require the presence of three or more macules to count as a major diagnostic feature3 and this was fulfilled in 65% of cases. A clinic based study19 reported hypopigmented macules in 90% of children with TS below 2 years of age and in 97% of children with TS overall. The frequencies of facial angiofibromas (75%), forehead plaques (19%), shagreen patches (48%) and ungual fibromas (15%) were also higher than reported here. This can in part be attributed to the different study design and the inclusion in this report of cases with a molecular genetic diagnosis of TS but not a definite clinical diagnosis. For facial angiofibromas, forehead plaques, shagreen patches and ungual fibromas, which increase in frequency through childhood and adolescence, the higher proportion of young children in our study is another important factor.
Data on eye findings were obtained for only 30 children in the study, probably because of incomplete capture of information on children referred for an ophthalmological assessment. It is possible this has biased the results. Retinal hamartomas were found in 40% of children. A clinic based study reported retinal lesions in 12% of children under 5 years of age and in 19% of children overall.17 A population based study found retinal hamartomas in 44% of TS patients.20
For 123 of the children (98%), the Vineland Adaptive Behaviour Scale and/or the Mullen Scales of Early Learning were completed and 65% showed some degree of intellectual disability (IQ<70). Studies of the frequency of intellectual disability in TS are prone to ascertainment bias. Population based studies provide the best data and have given estimates ranging from 44% to 65%.18 The young age of most of the children in the current study precluded an assessment of behaviour, which is the focus of ongoing studies.
Guidelines recommend that newly diagnosed cases of TS have renal ultrasound scanning for early detection of polycystic kidney disease (PKD) associated with contiguous gene deletions spanning the TSC2/PKD1 genes.1 4 5 In the present study 86% of cases had a renal ultrasound scan. Three cases of PKD were identified and molecular genetic testing confirmed a TSC2/PKD1 deletion in all of them. This demonstrates the importance of renal imaging in the initial diagnostic work-up. This condition causes severe early onset polycystic disease with progression to end-stage renal failure by early adult life.2 PKD was considered unlikely in one child who had multiple small cysts in both kidneys when first scanned at 5 months of age and confirmed at follow-up scans because there was no renal enlargement and no evidence of a TSC2/PKD1 deletion. Two other children had scans in early infancy reported as showing multiple echogenic lesions but subsequent scans were normal. These cases illustrate the caution needed in interpreting ultrasound findings in infancy. Renal ultrasound scanning demonstrated angiomyolipomas in 9% of children under 5 years of age and in 25% of older children. These figures are lower than those reported in a clinic based study that found angiomyolipomas in 17% of children under 2 years of age and an increasing frequency through childhood reaching 65% in cases aged 9–14 years.17 For renal cysts the frequency was 8% in the youngest children, rising to 12% at age 9–14 years, not so different from our finding of 7% overall.
Cardiac rhabdomyomas were detected in 61% of children under 5 years of age and in 36% of older children. Jóźwiak et al found similar frequencies of 61% in children under 5 years of age and 31% in older children.17 Other studies have reported rhabdomyomas in 50%–67% of children.21,–,23 In the cases with cardiac rhabdomyomas in our study, 40% of ECGs showed an abnormality, whereas no abnormal ECGs were reported in cases with normal echocardiography. It is notable that only a minority of children had an ECG, an investigation which is recommended in newly diagnosed cases because of the association with cardiac conduction defects and arrhythmia.4
MRI detected subependymal nodules and/or cortical tubers in 91% of children scanned. This is comparable to previous studies.17 24 As expected, the detection rate was higher than for CT scanning because of the better visualisation of cortical tubers. Five children had subependymal giant cell astrocytomas, including one case presenting prenatally, which is a rare occurrence but has been reported previously.13 25
In this study 30 children (24%) presented before the onset of seizures. Twenty-three of these cases subsequently had genetic testing which identified a mutation, including 10 cases who did not have a confirmed clinical diagnosis when they had their first seizure, demonstrating the contribution genetic testing can make in this situation. Given evidence that seizures, particularly in young children, may contribute to cognitive impairment,18 26,–,31 this raises important issues about monitoring for seizures and the indications for starting anticonvulsant therapy. Prospective EEG monitoring in young infants with TS has shown that the EEG often becomes abnormal with multifocal epileptic activity prior to the onset of clinically apparent seizures and there is evidence suggesting that these changes can be reversed by treatment with vigabatrin,32 33 giving better control of epilepsy and improving the cognitive and behavioural outcome.34 Further studies are needed to confirm that such early intervention does indeed prevent the development of epilepsy or reduce its severity, and crucially whether this has a beneficial effect on cognitive outcome.
In summary, diagnosis of TS can be challenging, particularly in young children, and this study has shown the value of genetic testing for confirming the diagnosis. Presentation and diagnosis before onset of seizures is becoming increasingly common, and raises difficult questions about whether these children should have EEG monitoring and concerning the criteria for starting anticonvulsant therapy.
The authors gratefully thank the many clinicians who helped with recruitment and responded to requests for clinical information; R Treacy, S Waller and L Selwood for retrieving laboratory reports; and all the patients and their families who kindly participated in the study.
Funding The authors thank the Tuberous Sclerosis Association and The Isaac Newton Trust for funding this study. PB is supported by the UK NIHR Biomedical Research Centre for Mental Health at the Institute of Psychiatry, King's College London and The South London and Maudsley NHS Foundation Trust.
Competing interests None.
Ethics approval The study was approved by West Midlands Multi-Centre Research Ethics Committee and by local research ethics committees for the participating centres.
Provenance and peer review Not commissioned; externally peer reviewed.
Members of the Tuberous Sclerosis 2000 Study Group not individually listed in the authorship are as follows: V Attard, A Clarke, FV Elmslie, AK Saggar, St George's Hospital, London; D Baines, BA Kerr, Royal Manchester Children's Hospital, Manchester; C Brayne, Institute of Public Health, University of Cambridge; C Connolly, A Lydon, C Srivastava, Institute of Psychiatry, King's College London; JA Cook, Sheffield Children's Hospital, Sheffield; C Falconer, St James's University Hospital, Leeds; DM Davies, Institute of Medical Genetics, Cardiff; AE Fryer, Alder Hey Children's Hospital, Liverpool; M Haslop, Y Granader*, University of Cambridge (*currently Yeshiva University, New York); PD Griffiths, University of Sheffield; A Hunt, Tuberous Sclerosis Association; WWK Lam, Western General Hospital, Edinburgh; JC Kingswood, Royal Sussex County Hospital, Brighton; ZH Miedzybrodzka, College of Life Sciences and Medicine, Aberdeen; H Crawford, PJ Morrison, Belfast City Hospital; BGR Neville, UCL Institute of Child Health, London; FJK O'Callaghan, University of Bristol; SG Philip, Birmingham Children's Hospital, Birmingham; S Seri, Aston Brain Centre, School of Life and Health Sciences, Aston University, Birmingham; R Sheehan-Dare, The General Infirmary, Leeds; CH Shepherd, Craigavon Area Hospital, Craigavon.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.