Background Benign hereditary chorea (BHC) is a rare autosomal dominant disorder characterised by childhood onset that tends to improve in adulthood. The associated gene, NKX2-1 (previously called TITF1), is essential for organogenesis of the basal ganglia, thyroid and lungs. The aim of the study was to refine the movement disorders phenotype. We also studied disease course and response to therapy in a large series of genetically proven patients.
Methods We analysed clinical, genetic findings and follow-up data in 28 NKX2-1 mutated BHC patients from 13 families.
Results All patients had private mutations, including seven new mutations, three previously reported mutations and three sporadic deletions encompassing the NKX2-1 gene. Hypotonia and chorea were present in early infancy, with delayed walking ability (25/28); dystonia, myoclonus and tics were often associated. Attention deficit hyperactivity disorder (ADHD) was present in seven. Among the 14 patients followed-up until adulthood, nine had persistent mild chorea, two had near total resolution of chorea but persistent disabling prominent myoclonus and three recovered completely. Learning difficulties were observed in 20/28 patients, and three had mental retardation. Various combinations of BHC, thyroid (67%) and lung (46%) features were noted. We found no genotype–phenotype correlation. A rapid and sustained beneficial effect on chorea was obtained in 5/8 patients treated with tetrabenazine.
Conclusion Early onset chorea preceded by hypotonia is suggestive of BHC. Associated thyroid or respiratory disorders further support the diagnosis and call for genetic studies. Tetrabenazine may be an interesting option to treat disabling chorea.
Statistics from Altmetric.com
Benign hereditary chorea (BHC) is a rare autosomal dominant movement disorder, defined by early onset in childhood, a stable or non-progressive course of chorea, and no mental deterioration. Mutations in the TITF1 gene (now named NKX2-1), located in chromosome region 14q13 and encoding the thyroid transcription factor1, have been identified.1 To date, about 30 different mutations have been reported.
Since its original description, the phenotype has been expanded to a combination of chorea, hypothyroidism and lung disease2–4 termed ‘brain–thyroid–lung syndrome’. NKX2-1 has three exons and plays a critical role during the embryogenesis of the thyroid, basal ganglia and lungs.5 Most papers on the subject report single cases or small series with various phenotypes and there are no systematic studies of BHC associated with TITF1/NKX2-1 mutations. Long term follow-up data are scarce.6 We therefore present the largest series of patients with genetically proven BHC to refine the disease spectrum and long term outcome. We comprehensively characterised the movement disorders and other neurological manifestations, evaluated treatment efficacy and assessed associations with lung and thyroid disorders.
Patients and methods
Referral centres in charge of molecular analysis of the NKX2-1 gene (Lille, Lyon and Paris) were contacted to gather data on BHC patients. Twenty-eight children and adults with genetically proven BHC were identified. All patients were examined and followed-up by at least one of the authors until 2011. In addition, all familial cases and 6/8 sporadic patients were clinically assessed using a standardised protocol (described below), either by physical examination or on video by three paediatric and adult movement disorders specialists (DD, ER, MV). The same assessment was performed on two sporadic patients that we could not examine personally. Clinical information was systematically collected from infancy or early childhood until the last examination in 2011. We analysed the following features: muscle tone in early childhood, developmental milestones and age at which they were reached, head circumference, school performance and cognitive function, behavioural disorders, initial severity and outcome of chorea, other abnormal movements, ataxia and gait disorders, functional disability (handwriting, gait and balance), effect of treatments (none, moderate, marked, disappearance of symptoms), follow-up duration at study endpoint (2011), and occurrence and characteristics of lung and thyroid disorders.
We then compared the clinical and genetic features of this series with those of patients previously reported in the literature. The NIH Pubmed database was screened to perform a systematic review of patients with NKX2-1 mutations reported since 1998, using the keywords ‘TITF1’, ‘NKX2-1’, and ‘benign hereditary chorea’.
Mutation analysis: DNA sequencing of the NKX2-1 gene
Blood was obtained with parents' informed consent for leucocyte genomic DNA extraction. Four pairs of NKX2-1 specific primers were used for PCR amplification of the coding region and the intron–exon boundaries (primers sequence available on request). The PCR products were subjected to direct sequencing in both directions using the Big Dye Terminator v3.1 kit (Applied Biosystems, Foster City, California, USA) and an ABI 3730 sequencer (Applied Biosystems). The reference sequence and exon numbering are those of Genbank accession No NM_003317.3.
This exploration was performed when the clinical phenotype was evoked with negative NKX2-1 mutation analysis. Genomic DNA was extracted from peripheral blood lymphocytes using the Flexigene DNA kit (Qiagen, Courtaboeuf, France).
In patients 25, 27 and 28, oligonucleotide array-CGH was performed using the Agilent Human Genome CGH 180K Microarray. This high resolution 60-mer oligonucleotide based microarray contains ∼170 000 probes spanning coding and non-coding genomic sequences with a median spacing of 13 kb. The arrays were analysed with the Agilent scanner (G2656BA). Patient CGH files extracted from Feature Extraction (Agilent Technologies, Santa Clara, CA, USA) were uploaded into Nexus Copy Number software V.5.1 (Biodiscovery Inc, El Segundo, California, USA).
We used the patient trio hybridisation strategy (DNA from three sex matched patients is hybridised on three different arrays: patient A vs patient B on array 1, patient C vs patient B on array 2 and patient A vs patient C on array 3. All deletions were mapped according to the March 2006 assembly of the UCSC genome browser (NCBI Build 36.1/hg18; http://genome.ucsc.edu).
A girl (family A, case No 1) was hypotonic during the first months of life and developed chorea by age 2 years. Motor milestones were reached as follows: sitting up at age 16 months, walking at age 2.5 years and normal language skills at age 4 years (with mild dysarthria). Gait was impaired, with frequent falls and balancing difficulties. Skilled hand movements were impaired by jerky movements. Neurological findings were otherwise normal. Head circumference was normal (0 SD). Thyroid function at age 3 years was normal. She was treated with fluticasone for asthma. Chorea was treated with tetrabenazine (0.5 mg/kg/day) from age 5 years, with a marked improvement as walking became almost normal and falls less frequent. Daily dosage was increased to 1 mg/kg/day but she developed asthenia and tetrabenazine was withdrawn. Chorea reappeared and tetrabenazine was reintroduced at a lower dose (0.7 mg/kg/day) with a subsequent marked improvement and good tolerance. At the last follow-up in 2011 (age 6 years), she still had subtle difficulties in handwriting and speech and required educational support and rehabilitation. She otherwise had a normal school record. Neuropsychological evaluation at age 6 years showed attention deficit hyperactivity disorder (ADHD) and a significant difference between her performance IQ (normal 103) and her verbal IQ (lower normal range 85). Genetic analysis revealed a c.257dupA mutation in the NKX2-1 gene, as in other affected members of her family.
Clinical analysis of our series and review of the literature
We identified 28 patients, consisting of eight sporadic cases and 20 members of five families in a retrospective study of all BHC patients referred to the centres in charge of the molecular analysis of NKX2-1. All patients originated from Europe, except for three who originated from Sri Lanka. Each family had a private mutation. We discovered seven new mutations in TITF1/NKX2-1 and three complete deletions, identified by array-CGH. Three of the 28 patients have previously been reported7 ,8 but were included in this survey as further outcome data were obtained up to 2011 in all three cases. Clinical and genetic characteristics of our series are summarised in table 1 (clinical features of brain, thyroid and lung disorders) and in supplementary table 2 (detailed neurological features; supplementary table 2 is available online only).
We compared these 28 patients with the 124 patients (32 families) from 29 papers described in the literature.1–3 9–31 Neurological, thyroid or lung status was missing for half of the patients described in the literature. Girls were more frequently affected than boys with a sex ratio of 0.64 in our series and 0.70 in the literature.
The spectrum of initial neurological manifestations was homogeneous in our series, with hypotonia in the first year of life, sometimes associated with delayed motor milestones, and early chorea. The chorea always improved up to puberty or early adulthood, and then remained stable until the sixth decade (longest follow-up 64 years). Among the 14 adults, five had complete (case numbers 5, 6 and 12) or near complete (case numbers 9 and 22) resolution of chorea, while nine had persistent mild chorea. Myoclonus (observed in eight cases) was intermingled with chorea in the early stages of the disease and became the main source of disability in adulthood (cases 7 and 9). Additional movement disorders included dystonia (cases 2, 8, 15, 19, 20, 22 and 23), and motor and vocal tics (cases 13, 14 and 28). Learning difficulties were observed in 20/28 patients. Quantified evaluation of cognitive performance performed in 14/20 showed that three had mental retardation, three a borderline IQ (between 70 and 80) and eight a normal IQ (supplementary table 2, available online only). The six other patients were examined at adult age and their learning disabilities during schooling were reported but not quantified. ADHD was observed in seven patients (six with normal IQ), two of whom improved on methylphenidate.
Mean follow-up was 24.5 years (range 4–64) in our series but could not be calculated for published cases. Among the eight children or adults who received tetrabenazine (mean 1 year, range 0.1–2.5 years), five rapidly experienced a moderate to marked benefit. Adverse effects were observed in four cases but were resolved by reducing the daily dosage in one patient (a child), without undermining the beneficial effect on chorea, while the other three patients (two children and one adult) stopped taking the treatment within days before efficacy could be evaluated (supplementary table 2, available online only). The literature contains only one description of successful treatment with tetrabenazine and one case of poor tolerability. In our series, three patients tried levodopa with moderate improvement in two and no benefit in one. Levodopa efficacy has been occasionally reported in the literature (n=3).14 ,29 Anecdotal reports have mentioned the possible efficacy of others drugs such as methylphenidate.18
Various combinations of BHC, lung and thyroid features (congenital hypothyroidism, compensated or patent hypothyroidism) were observed in our series, with ‘brain–thyroid–lung’ syndrome in 36% of cases (vs 49% in the literature), brain–thyroid disorders in 32% (vs 28%), brain–lung disorders in 10% (vs 1%) and isolated brain involvement (BHC) in 21% (vs 22%). For pulmonary complications, neonatal respiratory failure due to surfactant metabolism disorder was observed only once in our series (case No 23) but was more often mentioned in the literature (17 reported cases, including three deaths related to respiratory failure in infancy or later). Recurrent infections and asthma were mentioned throughout life, and lung cancer occurred once in our series and in three reported cases (two deaths).
Results of NKX2-1 molecular analysis
Mutation analysis and array results
Array results are summarised in supplementary table 3 (available online only). Array-CGH found a large 14q13.3 interstitial deletion encompassing NKX2-1 and PAX9 and adjacent genes in three cases with negative mutation analysis (supplementary table 3, available online only). Deletion sizes varied from 0.338 to 13.8 Mb (figure 2). Parents of the three index cases with deletions were healthy and did not carry the deletion.
This large series of patients with genetically proven BHC represents the longest reported follow-up (mean 24.5 years, range 4–64) and includes a systematic and exhaustive analysis of the molecular and clinical phenotype with a focus on the phenomenology of movement disorders, their course and response to therapy. We found seven novel mutations and characterised three complete TITF1/NKX2-1 deletions identified by array-CGH. The typical phenotype was infancy onset chorea preceded by hypotonia. Chorea was mainly isolated but could be associated with myoclonus, dystonia or tics, and improved after puberty. Myoclonus occasionally became the most disabling motor manifestation in adults. A high prevalence of associated learning disabilities and ADHD in our series suggests they represent underrated non-motor neurological manifestations of the disease. Tetrabenazine was effective both in children and adults with a good benefit/tolerance ratio.
Neurological features: homogeneous chorea phenotype and broad spectrum of associated movement disorders
All patients in our series (vs 97% in the literature) had neurological manifestations, owing to the fact that they were recruited in neuropaediatric and movement disorder clinics. Neurological dysfunction was revealed in most patients by marked hypotonia, sometimes accompanied by motor delay, with a mean walking age of 2.5 years (15 months–5.5 years). One child aged 5 years is still unable to walk. Age of unassisted walking reported in the literature is similar (mean 3 years, range 18 months–7 years); young children are described as clumsy and ‘ataxic’, with delayed walking.
Chorea started in infancy or childhood, with a mean age of 31 months at diagnosis (8 months–7 years), in line with the literature (mean age 2 years, range 6 months–8 years). Generalised chorea was present in all patients and sometimes worsened during premenstrual periods and pregnancy. Severity and outcome were variable within a given family and from one family to another, although stabilisation or complete resolution was always observed in early adulthood. Patients then remained stable over time, as shown by follow-up until the sixth decade.
Associated movement disorders
Myoclonus (n=8), upper limb dystonia (n=7), and motor and vocal tics (n=3) were observed in 12/28 cases (some of the patients had association of several movement disorders). When present, myoclonus was intermingled with chorea early in the disease and became the main source of disability as the chorea improved during adulthood. Dystonia is rarely mentioned in the literature27 ,30 and only a recent case report30 described the changing phenomenology of BHC over time: as in our observations, the patient evolved from chorea to EMG proven myoclonus associated with mild dystonia. Regarding this overlap between BHC and myoclonus in autosomal dominant cases, especially in adult patients, the most difficult differential diagnosis is myoclonus dystonia related to epsilon–sarcoglycan gene mutations. Asmus et al attempted to identify clinical criteria differentiating BHC from genetically proven myoclonus dystonia and concluded that the main clues to BHC were early onset hypotonia, chorea in infancy and possible association with thyroid or lung disease.19 Finally, in addition to the ‘classical’ features of BHC, other movements disorders (not observed in our series) have been described—namely, tremor21 ,27 and ataxia.1 ,7 ,13 ,22 This broader clinical spectrum may explain diagnostic difficulties in some cases of BHC. Schrag et al obtained follow-up information on 11 families in which a diagnosis of BHC was suspected but not genetically proven. Mean follow-up was 6.5 years (0–20 years) but five patients had none: in nine families, the initial diagnosis of BHC was modified to another diagnosis such as myoclonus dystonia, Huntington's disease, ataxia telangiectasia or primary dystonia.32
Beneficial effect of tetrabenazine on chorea
Tetrabenazine acts on the CNS, depleting monoamines and serotonin from nerve terminals by inhibiting their transport to presynaptic vesicles. It has been approved as a symptomatic treatment for chorea in Huntington's disease, at doses up to 100 mg/day. To date, the optimal dosage for children is poorly documented.
This treatment was evaluated retrospectively in eight patients. We could not use any standardised rating scale but parents, and children if possible, gave their own general assessment and the physicians an objective rating of the movement disorder process.
We observed a moderate to marked beneficial effect on chorea and motor function within days in 5/8 children and in adults with low dose tetrabenazine (0.5 mg/kg/day for children and 37.5 mg/day in adults). Tolerability was good except in three patients who experienced nausea, drowsiness or hyperexcitability and stopped taking tetrabenazine before clinical benefit could be evaluated. In the other patients, a sustained benefit was observed during a mean follow-up of 1.5 years (0.5 months–2.5 years). Relapse occurred when treatment was interrupted in case No 1, further supporting the fact that chorea improvement was not spontaneous.
In only a small number of patients, these data suggest efficacy of this treatment, which will require further prospective studies to be confirmed.
Low dose tetrabenazine might allow young patients to invest less effort in controlling their movement disorders and to focus on their schoolwork. We recommend starting with a very low dose of tetrabenazine (0.5 mg/kg/day), increasing in 0.5 mg/kg increments at weekly intervals, depending on tolerability, up to a maximum of 4.5 mg/kg/day, without exceeding 100–150 mg/day.
Only two detailed cases of tetrabenazine treatment for BHC have been reported in the literature28: a woman improved with tetrabenazine 75 mg/day and treatment interruption led to worsening of chorea as well as sleep disorders and behavioural problems; in the other case, tetrabenazine was stopped because of side effects (insomnia and anxiety).
The term ‘benign chorea’ should be revisited in view of these patients' cognitive and behavioural disturbances (ADHD). Seventy-one per cent of patients in our series experienced learning difficulties. Quantified evaluation of cognitive performance in 14 children showed that 15% had mental retardation and 15% were borderline (supplementary table 2, available online only). All of these patients required educational support. In the literature, cognitive functions are rarely precisely described; nevertheless, mental retardation is reported in 13% and learning difficulties in 25% of patients. Poor school performances may be due to movement disorders (leading to writing difficulties), slowness, memory disorders and/or attention deficit, and require special support. In addition, hypothyroidism could in itself be a cause of learning disability and cognitive impairment if not promptly diagnosed and treated. It may bias cognitive impairment prevalence in BHC subjects. However, even patients with early treatment showed decreased cognitive performances, not improved by supplementation.
ADHD was relatively frequent in our series (7/28 cases) whereas it is only mentioned in two cases in the literature.17 ,25 ADHD has been linked to morphological and functional striatal abnormalities, especially in the anterior (associative) striatum, both in humans33 and in animal models.34 ,35 A bilateral decrease in striatal volume36 and functional abnormalities of the basal ganglia28 ,37 have been described in a few cases of BHC. Basal ganglia dysfunction may involve both the motor (chorea) and non-motor (ADHD) striatal networks. An association with ADHD has also been reported in other types of chorea, such as Sydenham chorea.38 Therefore, similar dysfunction of the motor and associative striatal networks might account for the movement (chorea) and behavioural disorders (ADHD) in BHC. This hypothesis is in keeping with the structural alterations of the motor and associative striatum: a loss of most NKX2-1 mediated striatal interneurons has been reported in a case of genetically proven BHC.15
Systemic manifestations: thyroid and lung disorders
As NKX2-1 is essential for lung, thyroid and basal ganglia organogenesis, various combinations of brain, thyroid and lung involvement are reported in the literature, from isolated BHC to the ‘brain–thyroid–lung’ syndrome. Isolated BHC represented 21% of cases in our series (22% in the literature), ‘Brain–thyroid–lung’ syndrome was less frequent in our series (36%) than in the literature (49%) but is more likely to be seen in endocrinology and pulmonary medicine departments.1 ,2
Hypothyroidism is frequent in patients with NKX2-1 mutations (67% in our series and 34% in the literature with missing data). Congenital hypothyroidism and compensated hypothyroidism are the main manifestations. Congenital hypothyroidism with elevated thyroid stimulating hormone is detected by neonatal screening and is thus compensated for by early treatment. Psychomotor retardation is frequently attributed to thyroid hormone deficiency. If a baby with thyroid hormone deficiency has a poor neurological outcome despite adequate hormone replacement, NKX2-1 mutations should be suspected.17 Compensated hypothyroidism (isolated thyroid stimulating hormone elevation with normal T4) may be detected late, in childhood or adulthood. These patients are initially asymptomatic and may subsequently develop hypothyroidism. Therefore, it is important to regularly follow-up thyroid function in BHC patients so as to initiate treatment with L-thyroxin in a timely manner.7 ,11
Respiratory disorders are less frequent than neurological and thyroid disorders (46% in our series and 49% of published cases). The main disorders are (a) neonatal respiratory failure in full term newborns (alveolar syndrome explained by surfactant deficiency due to NKX2-1 mutation 82), (b) recurrent pulmonary infections, (c) asthma and (d) lung cancer,4 ,21 as NKX2-1 seems to play a role in carcinogenesis.39
Overall, there were no genotype–phenotype correlations with phenotype variability within the same family. Relation between the different mutations and protein functionality is still unknown.
In our series, mutations were de novo in 62% of index cases, and autosomal dominant inheritance was documented in 38% of cases, a distribution similar to that reported in the literature. Most of the NKX2-1 mutations previously reported truncate the beginning of the protein homeodomain, resulting in haploinsufficiency, the mutant protein being unable to bind DNA and activate target genes.22 In our series, 8/10 mutations were located in exon 3, containing the homeodomain. Frameshift mutations have also been described.17 In three cases, a large variable interstitial deletion encompassing the NKX2-1 gene was identified by array-CGH (figure 2). Both large (case No 25) and small (case No 28) deletions have a relatively benign phenotype (chorea, ADHD, tics, normal dentition) whereas medium size deletions (case No 27) presented with dental agenesis, chorea and hypothyroidism. It is unclear whether the size of the deletions influences phenotype severity.21 Patients with deletions including NKX2-1 and PAX9 can have oligodontia,18 ,24 malabsorption syndrome, xerostomia and xerophtalmia.18 Severe developmental retardation and microcephaly associated with thyroid and lung dysfunction have been described in two patients with deletion of chromosome region 14q12-13.3.2
In patients presenting with benign hereditary chorea with NKX2-1 gene mutations, we have expanded the motor phenotype (additional movement disorders such as dystonia, myoclonus and tics) and reported behavioural disorders such as ADHD. In addition, on a long term follow-up into the sixth decade, we observed improvement in the disabling chorea with sometimes residual myoclonus. In children, a rapid and sustained beneficial effect on chorea was obtained in five out eight patients treated with tetrabenazine. When a patient has a typical BHC phenotype, but no identifiable NKX2-1 mutation, array-CGH must be used to look for a deletion, even if the chorea is isolated.
We thank all of the patients and their families for their cooperation and support. We are grateful to Christine Raybaud, MD, from Pédiatric endocrinologic department in Hospices Civils de Lyon, France, for referring two patients, and David Young for editorial assistance.
Linked article 302834.
MV and DD contributed equally to this work.
Competing interests None.
Patient consent Obtained.
Ethics approval This was a retrospective study in which the data were collected during usual clinical care of the patients.
Provenance and peer review Not commissioned; externally peer reviewed.
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.