Long-term follow-up of type 1 lissencephaly: survival is related to neuroimaging abnormalities

Transcription

1 DEVELOPMENTAL MEDICINE & CHILD NEUROLOGY ORIGINAL ARTICLE Long-term follow-up of type 1 lissencephaly: survival is related to neuroimaging abnormalities MARIE-CLAIRE Y DE WIT 1 JOJANNEKE DE RIJK-VAN ANDEL 2 DICKY J HALLEY 3 PINO J PODDIGHE 3 WILLEMFRANSMARTS 1 IRENAEUS FM DE COO 1 GRAZIA MS MANCINI 3 1 Department of Neurology and Paediatric Neurology, Erasmus MC, Sophia Children's Hospital, Rotterdam, The Netherlands. 2 Department of Neurology, Amphia Hospital, Breda, The Netherlands. 3 Department of Clinical Genetics, Erasmus MC, Rotterdam, The Netherlands. Correspondence to Dr Grazia MS Mancini at Department of Clinical Genetics, Erasmus MC, PO Box 2040, 3000 CA Rotterdam, The Netherlands. g.mancini@erasmusmc.nl PUBLICATION DATA Accepted for publication 11th January Published online 17th March AIM To evaluate survival, clinical, and genetic characteristics of all patients with classic or type 1 lissencephaly born between 1972 and 1990 in the Netherlands, who at the time were enrolled in an observational study. METHOD We re-evaluated 24 patients (11 males, 13 females) for long-term follow-up and survival information. RESULTS Mean length of follow-up was 14 years (SD 9y 8mo). Eleven patients were alive at follow-up. All patients showed severe intellectual disability, intractable epilepsy, and complete dependency on care. Life expectancy was related to the severity of the lissencephaly on neuroimaging. Molecular analysis of the LIS1 gene was not possible at the time of the original study and was now requested by eight parents. This revealed a pathogenic nonsense mutation or deletion in seven patients. INTERPRETATION Our study provides information about the long-term course of lissencephaly and the relationship between lissencephaly severity and prognosis. It also shows that renewed attention to genetic counselling remains valued by families of patients with a severe congenital neurological disease. Lissencephaly ( smooth brain ) has long been recognized as a cause of intellectual disability. In type 1 lissencephaly the cerebral cortex lacks gyri and sulci and is thickened, either throughout the cortex or more in the posterior than the anterior regions. Early chromosome studies found a (partial) monosomy of chromosome 17 in some patients with lissencephaly with dysmorphic facial features (Miller Dieker syndrome). 1 In 1989, the identification the 17p13.3 region involved in Miller Dieker syndrome enabled prenatal testing in some cases. 2 Ten years later the LIS1 gene (also called the PAFAH1B1 gene) was identified in this region. 3 Several other genetic causes were later found in the lissencephaly, pachygyria, and subcortical band heterotopia spectrum, including the genes DCX, RELN, VLDLR, ARX, and TUBA1A. Most cases of type 1 lissencephaly are explained by mutations in LIS1 or by 17p13.3 haploinsufficiency. Other causes are sporadic, although the p.r402c mutation in the TUBA1A gene has been reported in several cases, but in less than 1% of patients with lissencephaly type 1 with cerebellar hypoplasia. 4 Clinical signs are epilepsy, intellectual disability, and spasticity. Not much has been published on the follow-up of patients with lissencephaly or their life expectancy. Groups of patients are describedingeneraltermsandfollow-upisreportedtoearly childhood and, in a few cases, until adulthood. 5 9 To describe the natural course of lissencephaly type 1 and improve prognostic information, we re-evaluated all patients with type 1 lissencephaly identified in a nationwide study conducted in the Netherlands between 1972 and 1990 who, if alive, would now be adults. 7,10 12 This is the largest reported cohort of consecutive patients. At the time of the original study, parents had given informed consent for DNA to be stored and to be contacted if and when further genetic testing would become feasible. METHOD Twenty-four Dutch patients born between 1972 and 1990 with type 1 lissencephaly were included in the original study. 7,10 12 At the end of the original follow-up, 17 patients were alive and we attempted to contact these. If possible, patients were examined at our clinic, otherwise a telephone interview was conducted with the parents caregivers. All original brain imaging (computed tomography [CT] and magnetic resonance imaging [MRI]) was re-evaluated and graded according to the following lissencephaly patterning scale: grade 1, complete agyria; grade 2, agyria with few shallow anterior sulci; grade 3, posterior agyria and anterior pachygyria; grade 4, diffuse pachygyria posterior>anterior; grade 5, mixed pachygyria and subcortical band heterotopia; grade 6, subcortical band heterotopia only. 13 If the patient had died, age and cause of death were recorded. During the original study, DNA from leukocytes of patients alive at diagnosis ª The Authors. Developmental Medicine & Child Neurology ª 2011 Mac Keith Press DOI: /j x 417

2 (18 24) was stored with the parents informed consent. If genetic analysis had not previously been carried out, we offered the parents testing for 17p13.3 deletion, sequence analysis of LIS1 and DCX where appropriate for the phenotype, and multiplex ligation-dependent probe amplification (MLPA) analysis for deletions in these genes. Repeating neuroimaging or electroencephalography (EEG) was not considered justified as it would not benefit the patients. RESULTS Diagnosis In two patients lissencephaly had been diagnosed at autopsy, for all others neuroimaging was available. Lissencephaly grading was based on the autopsy report or CT of the brain (Fig. 1). Additional MRI was available for five patients showing the same lissencephaly grade as on CT. Most had a grade 1 or 2 lissencephaly (eight and 11 patients respectively). Five patients had a grade 3 or 4 lissencephaly (Table I). Grade 1 Grade 2 Grade 3 Figure 1: Computed tomography (CT) scans and grading of lissencephaly. Grade 1 (patient 24, female, CT performed at age 9mo) shows complete agyria, `figure of eight' configuration due to lack of opercularization, and thickened cortex. Also note the cavum septum pellicidum. In grade 2 (patient 17, male, CT performed at age 4.5mo) note the few shallow sulci in the frontal lobes and some opercularization. In grade 3 (patient 4, male, CT performed at age 1y) note diffuse course gyri and shallow sulci predominantly in the anterior regions, and the further opercularization of the insular region. What this paper adds The survival of patients with type 1 lissencephaly is related to the severity grading on neuroimaging. Children with grade 1 to 2 lissencephaly all have severe psychomotor motor impairment and intractable epilepsy. Our findings support the relationship of a severe phenotype with nonsense mutations or intragenic deletions of LIS1. Survival During the original study period seven out of 24 patients died at ages between 0 and 9 years (Table I and Fig. 2). Parents caregivers of all other patients agreed to visit the clinic or give information by telephone, except for one who was lost to follow-up (patient 5). Five patients had died since 1990, aged 3 to 19 years. Cause of death was status epilepticus or pneumonia. One patient had been recorded as deceased, but we could not reliably recover the age of death (patient 5). Eleven patients are alive, now aged 18 to 27 years (Table I and Fig. 2). Five visited the outpatient clinic. Survival is associated with the severity of lissencephaly. The relation between survival and lissencephaly remained present when only patients with a molecularly confirmed LIS1 mutation were analysed. The overall survival would be overestimated by looking only at the molecularly confirmed patients, as genetic analysis was performed more often in patients who survived longer. There was also no DNA available for children who were diagnosed by autopsy. Clinical signs All patients showed severe psychomotor disability (Table I). Their IQ was too low for formal testing and was assessed into three categories: (1) most severe: no eye contact or purposeful movements sounds (nine patients); (2) severe: some signs of recognition of parents and stimulus-related laughing (14 patients); (3) less severe: some purposeful movements or sounds (one patient). In one patient the ability to interact was lost at the onset of seizures. All had intractable epilepsy with age at onset ranging from a few days to 2 years (mean 6mo). Five patients presented with neonatal seizures, 17 with infantile spasms, and two with multifocal epilepsy. There was no correlation between age at onset of seizures and survival. In two patients seizures were reasonably controlled by antiepileptic drugs; the others had daily to weekly seizures. The EEGs have been described before and show typical generalized fast activity with high amplitude and or high amplitude sharpand slow-wave complexes in almost all cases. 12 No patient ever had a normal EEG. Two patients had shown some motor development to head balance, clapping hands, and belly crawling, but had lost these abilities during periods of infectious disease and seizures. None was ever ambulant. All patients show axial hypotonia, and four patients have a severe scoliosis. Spasticity was mild, but more prominent in three patients. All but one of the parents said that having a child with lissencephaly had a severe and negative impact on their family and their quality of life, but that their child appeared to be content most of the time. One mother stated that she felt that her child did not have any quality of life. Genetic analysis During the original study, microscopic high-resolution chromosome banding was carried out for all patients; in one patient this had shown a probable chromosome 17p13.3 deletion consistent with Miller Dieker lissencephaly syndrome (patient 7). This was confirmed with fluorescence in situ hybridization (FISH) analysis. This patient died during the original study period. Since 1990 three more patients had been diagnosed with Miller Dieker syndrome due to a deletion at chromosome 17p13.3 (patients 13, 20, and 24). One of these has since died. DNA was still stored for 14 patients without a molecularly confirmed diagnosis. A total of six patients had died after 1990, five without a molecularly confirmed diagnosis. Parents of one of these wanted further testing to aid reproductive choices in their other children: a heterozygous LIS1 mutation (c.334dupa) 418 Developmental Medicine & Child Neurology 2011, 53:

3 Table I: Original cohort: final diagnosis at follow-up No. Sex Lissencephaly grade Genetics Age (y) Age at death Estimated best motor developmental level (mo) ID Age at epilepsy onset Epilepsy type Chewing swallowing Cause of death (where applicable) 1 M 2 (CT, MRI) LIS1 c. 569-insTAA y Multifocal seizures Gastrostomy 2 M 2 (CT) NA 6mo wk Neonatal seizures NA Unknown 3 F 3(CT) LIS1 c.900+1g>c mo Infantile spasms Gastrostomy 4 M 3 (CT, MRI) LIS1 c.162dela mo Infantile spasms F 1(CT) LIS1, MLPA negative Lost to follow-up mo Infantile spasms NA Unknown age 5y 6 F 2 (CT) Refused mo Infantile spasms ++ 7 F 1 (CT) 17p13.3 deletion 3y mo Infantile spasms NA Unknown 8 F 3(CT) LIS1, TUBA1A, MLPA negative mo Infantile spasms + 9 M 1 (autopsy) NA 9y mo Infantile spasms NA Status epilepticus 10 F 1 (autopsy) NA 3y mo Infantile spasms NA Pneumonia 11 F 3 (CT) LIS1 c.782g>a mo Neonatal seizures + 12 F 1 (CT) Refused 19y mo Infantile spasms Gastrostomy Unknown 13 M 2 (CT, MRI) 17p13.3 del (FISH) mo Infantile spasms NA 14 F 1 (CT and autopsy) NA 1mo wk Neonatal seizures NA Pneumonia 15 F 1 (CT and autopsy) NA 7mo mo Infantile spasms NA Pneumonia 16 F 2 (CT and autopsy) NA 6y mo Infantile spasms NA Unknown 17 M 2 (CT) Refused 12y mo Infantile spasms + Pneumonia 18 M 4 (MRI) Refused mo Infantile spasms M 2 (CT) Refused 16y mo Multifocal epilepsy +++ Pneumonia 20 F 2 (CT) 17p13.3 deletion mo Neonatal seizures + 21 M 2 (CT, MRI) LIS1 deletion (MLPA) mo Infantile spasms M 2 (CT) LIS1 del exon 6 11 (MLPA) mo Infantile spasms + 23 M 2 (CT, MRI) LIS1 c.334dupa 20y 1 ++ Intractable epilepsy Gastrostomy Pneumonia 24 F 1 (CT) 17p13.3 deletion 2y wk Neonatal seizures Gastrostomy Heart failure M, male; F, female; CT, computed tomography; MRI, magnetic resonance imaging; MLPA, multiplex ligation-dependent probe amplification (lissencephaly kit MRC Holland, Amsterdam, the Netherlands, containing probes for each of the LIS1 exons and including probes for the DCX, POMT1, POMGnT1, and FLNA genes); FISH, fluorescence in situ hybridization; NA, not available; ID, intellectual disability: +++, no contact; ++, little contact and laughing; +, some play, no language; Chewing swallowing: +, fed formula or pureed food without problems; ++, mild dysphagia; +++, dysphagia with recurrent pneumonia. Long-term Follow-up of Lissencephaly Marie-Claire Y de Wit et al. 419

4 Original cohort n=24 Genetic analysis Died n=7 6 no DNA 1 17p13.3 deletion Follow up 2008 Lost to follow-up n=1 LIS1 excluded Re-evaluated n=16 Since died n=5 1 17p13.3 deletion 1 LIS1 mutation 3 refused Alive n=11 4 LIS1 mutation 2 MLPA del LIS1-2 17p13.3 deletion 1 LIS1 excluded 2 refused Figure 2: Flow chart of follow-up results. was detected (patient 23). Parents of two patients preferred to leave the decision to their other children; one couple did not have other children. One family was lost to follow-up. Eleven patients were still alive. In two a chromosome 17p13.3 had been diagnosed after 1990 (patients 13 and 20, see above). Of the remaining nine patients, seven families requested further genetic testing to confirm the diagnosis and to aid reproductive choices in the patients siblings. Two families preferred not to have any more medical tests. Direct sequence analysis of the coding exons and exon intron boundaries showed heterozygosity for a frameshift, nonsense, or splice site mutation in the LIS1 gene in four patients (see Table I). MLPA analysis showed an intragenic LIS1 deletion in one patient spanning at least the six most 3 exons, whereas a second patient had a total LIS1 deletion possibly further extending into the Miller Dieker region. In one patient, LIS1 sequencing and MLPA testing of the LIS1 gene were normal (patient 8). Additional sequence analysis of TUBA1A was also normal. When a gene defect was identified, parents and siblings were invited for genetic counselling. In parents who were tested, mutations were absent (two couples). When mutations were found, most parents were pleased that any doubts about influences during pregnancy and risks of recurrence had now been dispelled. DISCUSSION Our cohort study confirms that children with severe grades of type 1 lissencephaly have a severe intellectual and motor disability and that the epilepsy is intractable in all cases. Treatment for epilepsy is still important, as seizures can lead to loss of skills or death as demonstrated in this cohort. Life expectancy is limited, but with supportive care focused on the prevention of infectious complications and scoliosis, many of these children do reach adulthood. Survival may be related to the quality of care, for example, many of the younger patients died before the era of routine gastrostomy (see Table I). It will be interesting to see if survival changes in cohorts spanning the next two decades. All patients in this cohort have severe lissencephaly phenotypes. The lissencephaly phenotype can also be mild, for example with epilepsy and normal cognitive development. 14 Milder phenotypes, such as subcortical band heterotopia, were often not recognized during life before 1990 owing to limitations of neuroimaging. We show that life expectancy is related to the severity of the lissencephaly on neuroimaging. This relationship remains when only those patients with molecularly confirmed LIS1 mutations are included. We chose not to exclude non-molecularly confirmed cases beforehand as that would have biased the results towards overestimating survival. In contrast to some other types of cortical malformation, type 1 lissencephaly shows a typical and homogeneous pattern and probably reflects the same pathological mechanism, even when different genetic causes are involved. 4 Location of the mutations in the LIS1 gene has been postulated not to predict disease severity directly. 15,16 However, LIS1 missense mutations are sometimes thought to result in a milder phenotype than nonsense mutations. 14 Also, the severity of the effect of the mutation on the LIS1 protein does have a relationship with the severity of the lissencephaly grading. 9,17,18 In our cohort all mutations found in LIS1 were nonsense mutations or intragenic deletion, and we did not find any missense mutations. There was no indication of 420 Developmental Medicine & Child Neurology 2011, 53:

5 somatic mosaicism in our cohort, although this can never be excluded. Possibly, our findings are caused by selection bias towards the more severe phenotype owing to the neuroimaging limitations, indirectly supporting the notion that missense mutations may result in milder phenotypes. With the techniques available at the time of the original study, only microscopic chromosome deletions could be detected by high-resolution chromosome banding: a deletion of 17p was found in one patient then (later confirmed by FISH analysis). After 1990 a deletion of chromosome 17p13.3 was found in three more patients. When we offered parents further genetic analysis, eight out of 14 couples welcomed this, and in seven patients genetic testing using current techniques confirmed the clinical diagnosis. These results confirm the high predictive value of type 1 lissencephaly on neuroimaging for LIS1 mutations or deletions. In one patient we were unable to confirm a genetic cause with all available options, including TUBA1A analysis (patient 8). Possibly she has a LIS1 dysfunction that could not be proven with our methods (MLPA and sequencing) or the available material (DNA isolated from leukocytes). Our study shows the importance of the interpretation and classification of neuroimaging for prognosis and counselling of parents of a child with a severe congenital neurological disease. It also reminds us that genetic studies are still valued by parents and siblings of adult disabled patients to understand the cause, and to answer any concerns about risk of recurrence. ACKNOWLEDGEMENTS Dr D T Pilz, Department of Medical Genetics, Cardiff University School of Medicine, UK, tested patient 8 for TUBA1A mutations. The research in this paper was supported by a grant from Erasmus MC. REFERENCES 1. Dobyns WB, Stratton RF, Parke JT, Greenberg F, Nussbaum RL, Ledbetter DH. Miller Dieker syndrome: lissencephaly and monosomy 17p. J Pediatr 1983; 102: Stratton RF, Dobyns WB, Airhart SD, Ledbetter DH. New chromosomal syndrome: Miller Dieker syndrome and monosomy 17p13. Hum Genet 1984; 67: Reiner O, Carrozzo R, Shen Y, et al. Isolation of a Miller Dieker lissencephaly gene containing G protein b-subunitlike repeats. Nature 1993; 364: Kumar RA, Pilz DT, Babatz TD, et al. TUBA1A mutations cause wide spectrum lissencephaly (smooth brain) and suggest that multiple neuronal migration pathways converge on alpha tubulins. Hum Mol Genet 2010; 19: Kurlemann G, Schuierer G, Kuchelmeister K, Kleine M, Weglage J, Palm DG. Lissencephaly syndromes: clinical aspects. Childs Nerv Syst 1993; 9: Gastaut H, Pinsard N, Raybaud C, Aicardi J, Zifkin B. Lissencephaly (agyria pachygyria): clinical findings and serial EEG studies. Dev Med Child Neurol 1987; 29: de Rijk-van Andel JF, Arts WFM, Hofman A, Staal A, Niermeijer MF. Epidemiology of lissencephaly type I. Neuroepidemiology 1991; 10: Dobyns WB, Reiner O, Carrozzo R, Ledbetter DH. Lissencephaly. A human brain malformation associated with deletion of the LIS1 gene located at chromosome 17p13. JAMA 1993; 270: Fogli A, Guerrini R, Moro F, et al. Intracellular levels of the LIS1 protein correlate with clinical and neuroradiological findings in patients with classical lissencephaly. Ann Neurol 1999; 45: de Rijk-van Andel JF, Arts WF, Barth PG, Loonen MC. Diagnostic features and clinical signs of 21 patients with lissencephaly type 1. Dev Med Child Neurol 1990; 32: de Rijk-van Andel JF, van der Knaap MS, Valk J, Arts WF. Neuroimaging in lissencephaly type I. Neuroradiology 1991; 33: de Rijk-van Andel JF, Arts WF, de Weerd AW. EEG and evoked potentials in a series of 21 patients with lissencephaly type I. Neuropediatrics 1992; 23: Kato M, Dobyns WB. Lissencephaly and the molecular basis of neuronal migration. Hum Mol Genet 2003; 12: R Leventer RJ, Cardoso C, Ledbetter DH, Dobyns WB. LIS1 missense mutations cause milder lissencephaly phenotypes including a child with normal IQ. Neurology 2001; 57: Uyanik G, Morris-Rosendahl DJ, Stiegler J, et al. Location and type of mutation in the LIS1 gene do not predict phenotypic severity. Neurology 2007; 69: Saillour Y, Carion N, Quelin C, et al. LIS1-related isolated lissencephaly: spectrum of mutations and relationships with malformation severity. Arch Neurol 2009; 66: Cardoso C, Leventer RJ, Matsumoto N, et al. The location and type of mutation predict malformation severity in isolated lissencephaly caused by abnormalities within the LIS1 gene. Hum Mol Genet 2000; 9: Cardoso C, Leventer RJ, Dowling JJ, et al. Clinical and molecular basis of classical lissencephaly: mutations in the LIS1 gene (PAFAH1B1). Hum Mutat 2002; 19: Long-term Follow-up of Lissencephaly Marie-Claire Y de Wit et al. 421