Article Text

Original research
Adult phenotype of KCNQ2 encephalopathy
  1. Stephanie Boets1,
  2. Katrine M Johannesen2,3,
  3. Anne Destree4,
  4. Filippo Manti5,
  5. Georgia Ramantani6,
  6. Gaetan Lesca7,8,
  7. Laurent Vercueil9,
  8. Mary Kay Koenig10,
  9. Pasquale Striano11,12,
  10. Rikke Steensbjerre Møller2,3,
  11. Edward Cooper13,
  12. Sarah Weckhuysen1,14,15
  1. 1 Neurology Department, University Hospital Antwerp, Antwerp, Belgium
  2. 2 Department of Epilepsy Genetics and Personalized Treatment, The Danish Epilepsy Centre Filadelfia, Dianalund, Denmark
  3. 3 Institute for Regional Health Research, University of Southern Denmark, Odense, Denmark
  4. 4 Department of Human Genetics, Institute of Pathology and Genetics, Gosselies, Belgium
  5. 5 Department of Human Neuroscience, University of Rome La Sapienza, Roma, Lazio, Italy
  6. 6 Department of Neuropediatrics, University Children's Hospital, Zurich, Switzerland
  7. 7 Department of Genetics, University Hospitals of Lyon, Lyon, France
  8. 8 Neuroscience Research Center, Claude Bernard Lyon I University, Lyon, France
  9. 9 Grenoble Institute of Neurosciences (GIN), University Grenoble Alpes, La Tronche, France
  10. 10 Department of Pediatrics, University of Texas McGovern Medical School, Houston, Texas, USA
  11. 11 Department of Neurosciences, Rehabilitation, Ophthalmology, Genetics, Maternal and Child Health, University of Genoa, "G Gaslini" Institute, Genova, Italy
  12. 12 Pediatric Neurology and Muscular Diseases Unit, IRCCS’ G Gaslini" Institute, Genova, Italy
  13. 13 Department of Neurology, Baylor College of Medicine, Houston, Texas, USA
  14. 14 Applied & Translational Neurogenomics Group, VIB-Center for Molecular Neurology, VIB, Antwerp, Belgium
  15. 15 Translational Neuroscience Group, University of Antwerp, Antwerp, Belgium
  1. Correspondence to Prof. Dr. Sarah Weckhuysen; sarah.weckhuysen{at}


Background Pathogenic KCNQ2 variants are a frequent cause of developmental and epileptic encephalopathy.

Methods We recruited 13 adults (between 18 years and 45 years of age) with KCNQ2 encephalopathy and reviewed their clinical, EEG, neuroimaging and treatment history.

Results While most patients had daily seizures at seizure onset, seizure frequency declined or remitted during childhood and adulthood. The most common seizure type was tonic seizures (early) infancy, and tonic-clonic and focal impaired awareness seizures later in life. Ten individuals (77%) were seizure-free at last follow-up. In 38% of the individuals, earlier periods of seizure freedom lasting a minimum of 2 years followed by seizure recurrence had occurred. Of the 10 seizure-free patients, 4 were receiving a single antiseizure medication (ASM, carbamazepine, lamotrigine or levetiracetam), and 2 had stopped taking ASM. Intellectual disability (ID) ranged from mild to profound, with the majority (54%) of individuals in the severe category. At last contact, six individuals (46%) remained unable to walk independently, six (46%) had limb spasticity and four (31%) tetraparesis/tetraplegia. Six (46%) remained non-verbal, 10 (77%) had autistic features/autism, 4 (31%) exhibited aggressive behaviour and 4 (31%) destructive behaviour with self-injury. Four patients had visual problems, thought to be related to prematurity in one. Sleep problems were seen in six (46%) individuals.

Conclusion Seizure frequency declines over the years and most patients are seizure-free in adulthood. Longer seizure-free periods followed by seizure recurrence are common during childhood and adolescence. Most adult patients have severe ID. Motor, language and behavioural problems are an issue of continuous concern.

  • epilepsy
  • prognosis
  • genetics
  • medical
  • phenotype
  • congenital
  • hereditary
  • and neonatal diseases and abnormalities

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information.

Statistics from

Request Permissions

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.


KCNQ2 encephalopathy is a severe epileptic disorder characterised by early (mostly neonatal) onset therapy-resistant seizures and various degrees of developmental delay, caused by pathogenic variants in the gene KCNQ2. KCNQ2 encephalopathy was shown to be the most frequent genetic cause of neonatal onset epileptic encephalopathy, with an estimated incidence of 3/100.000 births.1 2 KCNQ2 encodes for the Kv7.2 subunit of the voltage-gated potassium channel that is responsible for the M-current, an important regulator of neuronal excitability. Like other voltage-gated K+ channels, functional Kv7 channels are tetramers of subunits, each having six transmembrane segments (S1–S6) and cytoplasmic N-terminal and C-terminal of variable length.3–5 Heterozygous pathogenic variants in KCNQ2 are responsible for epileptic disorders with various modes of genetic inheritance and clinical severity.3 The difference in phenotype can be explained by the different functional effects on the M-current of the respective amino acid substitutions.6 7 Inherited KCNQ2 variants with pure loss-of-function effects are known to lead to autosomal dominant benign familial neonatal epilepsy (BFNE).8–11 Specific KCNQ2 missense variants give rise to KCNQ2 encephalopathy.6 12 13 These variants arise de novo or are inherited from a mosaic parent, and have either a dominant-negative14 or more rarely, a gain-of-function effect.12 Both syndromes start early within the first days of life. Patients diagnosed with BFNE have a good prognosis regarding seizure remission and neurodevelopment.8 Patients with KCNQ2 encephalopathy suffer from early onset refractory seizures and developmental delay of varying severity.6 15

Several studies describe the disease course of children with KCNQ2 encephalopathy, but so far, no studies on the outcome in adulthood exist. This impedes the counselling of parents of young children diagnosed with KCNQ2 encephalopathy about the long-term outcome with regards to seizure frequency, neurological impairments, cognitive functioning and behaviour.

We describe the adult phenotype of patients with a KCNQ2 encephalopathy to inform other parents of children with KCNQ2 encephalopathy about the future of their children.



This is a retrospective study of adults (age ≥18 years at inclusion) with neonatal or early infantile onset epilepsy, genetic testing showing a pathogenic KCNQ2 variant as defined by ACMG criteria,16 and clinical histories indicating some degree of neurodevelopmental delay. The presence of some degree of neurodevelopmental delay was an explicit inclusion criterion to exclude individuals with B(F)NE. Patients with additional (likely) pathogenic variants leading to neurodevelopmental disorders or epilepsy were excluded. Eligible subjects were identified from those enrolled in the RIKEE patient and variant registry ( by their parent or treating physician. Also, individuals fulfilling inclusion criteria were located by contacting collaborating (paediatric) neurologists and medical geneticists. All parents or legal caregivers of recruited patients provided informed consent for inclusion in our study.

Clinical data collection

Referring physicians were provided with a data sheet to collect clinical information. Personal and familial medical history, results of genetic testing, neurological examination, detailed epilepsy and developmental history, and information on EEG and imaging results were obtained. All information was extracted from existing medical files by the treating physicians. Seizure types were classified using International League Against Epilepsy terminology based on the description provided by the treating physician.17 Age at epilepsy improvement was defined as the age at which a clinically relevant decrease in seizure frequency was seen, as noted in the medical records of the treating physicians. Intellectual disability (ID) was defined based on the level of adaptive functioning as proposed by the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition.18


Demographic and genetic features

A total of 24 unrelated adult individuals with KCNQ2 encephalopathy were identified and clinical data collected. For 11, insufficient clinical data from adulthood could be retrieved, so they were excluded. The clinical features of the 13 individuals where extensive data could be collected are summarised in table 1 and described in detail in online supplemental table 1. They originated from Italy, France, Denmark, Belgium, Switzerland and the USA. Five of the 13 were male (38%) and 8 were female. The age of inclusion in this study ranged from 18 years to 45 years.

Table 1

Key clinical features of patients with KCNQ2 encephalopathy

Sequencing revealed 10 different heterozygous KCNQ2 missense variants in 12 patients, and an in-frame trinucleotide deletion in one (c.314_316delCCT, p.Ser105del; patient 1). Three individuals (patients 3, 4 and 5) carried the same missense variant c.619C>T, p.Arg207Trp. Seven missense variants were previously reported in (unrelated) younger patients diagnosed with KCNQ2 encephalopathy (table 1).6 13 15 19–36 Three missense variants (c.367G>A, p.Glu123Lys; c.629G>C, p.Arg210Pro; c. 1040A>G p.Tyr347Ser) are, to our knowledge, novel.

In 10 individuals the KCNQ2 variants arose de novo. Inheritance was unknown for two individuals (patients 4 and 11). The mother of patient 4, who also had a mild ID, declined further genetic testing. The variant identified in patient 11 was not found in the mother, but the father was not available for testing. The in-frame trinucleotide deletion was inherited from a mosaic mother with self-limiting neonatal seizures, a single seizure during adulthood and normal development (patient 1). This family has been reported in a genetic study on BFNE in 2004 (prior to description of KCNQ2 encephalopathy), but follow-up of the proband revealed clear neurodevelopmental delay.19 Seven individuals had a family history of seizures in one or more first-degree or second-degree relatives. The KCNQ2-variant segregated in two of these families (family of patient 1 and patient 5). Two individuals had a family history of ID: the mother of patient 4, who declined genetic testing, and the son of patient 5 who also had a phenotype compatible with KCNQ2 encephalopathy.

Evolution of seizure frequency

All but one individual (patient 5) had seizure onset within the first week of life. Age at seizure onset ranged from the first day of life to 2 months (median: second day).

Seizure frequency decreased with increasing age. At onset, all patients for whom this information was available had one or multiple seizures daily. The age at epilepsy improvement varied from the first month of life until the age of 8 years (median: 12 months). Six individuals (46%) had improvement during the first year of life, another four individuals had improvement between the age of 1 year and 2 years 6 months (31%). In 5 out of 13 individuals, seizure-free periods of a minimum of 2 years followed by seizure recurrence had occurred before inclusion in the study.

By the age of 18 years, 6 out of 13 individuals were seizure-free for at least 2 years (46%). At the time of inclusion in the study, eight individuals (62%) were between 2 years and 19 years (median 6 years) seizure-free with ongoing antiseizure medication (ASM) use, and two (15%) were seizure-free for 16 years and 20 years, respectively, and were no longer taking ASM (patient 7 and 8). Three individuals (patients 4, 11 and 12; 23%) had ongoing seizures in adulthood (figure 1). One individual (patient 11) still had several seizures daily. This individual was born prematurely and thought to have suffered perinatal asphyxia. No information on cerebral imaging was however available. The two other individuals had infrequent seizures.

Figure 1

Age at epilepsy offset. Y, years.

Evolution of seizure types

Seizure types were defined based on the description of the treating physician and available (inter)ictal EEG findings. The onset of tonic, clonic and tonic-clonic seizures (focal vs generalised) was often not clear from the description and available historical data, therefore onset type was not specified for these categories of seizures. Seizure types include focal impaired awareness, tonic, clonic, tonic-clonic, myoclonic, epileptic spasms and status epilepticus (figure 2). All individuals had several seizure types, which often changed throughout life. The most frequent types were tonic-clonic, focal impaired awareness and tonic seizures. Tonic seizures were prominent at onset and during infancy, but did not occur in adulthood. In adulthood, the most frequent seizures were focal impaired awareness seizures and tonic-clonic seizures (figure 2).

Figure 2

Distribution of seizure types over the different age categories. Y, years.

Antiseizure medication

As noted, 2 out of 10 seizure-free patients were receiving no ASM at the time of inclusion in our study. In 4 out of 10 seizure-free patients, seizures were controlled with carbamazepine, lamotrigine or levetiracetam monotherapy. The remaining four seizure-free patients used polytherapy in different combinations to control their seizures. In total, seven of the eight seizure-free patients still receiving ASM were being treated with a sodium channel blocker (SCB): carbamazepine or lamotrigine. One of the three patients who was not seizure-free in adulthood had never been treated with an SCB (patient 11). Of note, carbamazepine was said to have aggravated seizures in one individual (patient 4).

EEG and brain imaging

EEG findings at onset were available for nine individuals and included burst-suppression, discontinuous trace, focal and multifocal epileptiform activity. During childhood, EEG showed multifocal (poly)spikes or sharp waves, spike-wave complexes and slow background activity. Four out of the 10 patients with available EEG during adulthood had unremarkable findings.

Brain MRI was performed in 11 individuals at various ages and was normal in 6. In five individuals, non-specific abnormalities were seen, including enlarged ventricles (patient 8), decrease in white matter volume (patient 12) and severe frontal and insular cortical atrophy (patient 6), to bi-occipital ischaemic abnormalities and a minor thalamic haemorrhage (patient 4). Patient 10 was diagnosed with right parietal cortical dysplasia. The interictal EEGs at onset and during childhood showed multifocal interictal epileptiform activity and no ictal EEGs were available to support (or exclude) the presence of a seizure onset zone within the dysplastic cortex. Although we thus cannot exclude the role of this dysplastic tissue in seizure generation, the multifocal interictal EEG abnormalities suggest more diffuse brain hyperexcitability.

Neurological examination and development

ID ranged from mild to profound (figure 3). The largest subgroup had severe ID (7/13; 54%). Five individuals (38%) did not receive any type of school education, the other eight (62%) had special education.

Figure 3

Cognitive functioning in adulthood. ID, intellectual disability.

Six (46%) individuals did not develop speech and in four individuals (31%), speech was limited to a few words or sentences. Seven (54%) walked independently, four (31%) were wheelchair-bound and two (15%) could walk with assistance. All adult individuals needed assistance in daily life activities.

Motor abnormalities were prominent, including limb spasticity in six (46%) individuals, and tetraparesis or tetraplegia in four (31%). Other neurological impairments were axial hypotonia, pyramidal signs, dystonia, dysarthria, apraxia and ataxia.

Neuropsychiatric features

Ten individuals (77%) showed autistic features or were diagnosed with autism spectrum disorder (figure 4). Four (31%) exhibited aggressive behaviour, four (31%) showed destructive behaviour with self-injury and two (15%) had anxiety problems (patients 5 and 7). Patient 10 often needed to be immobilised in his wheelchair to protect him from self-injury during aggressive outbursts. Sleep problems during childhood and adulthood were seen in six (46%) individuals and were described by the treating physician as insomnia, nocturnal restlessness, unspecified uncontrolled movements, bruxism and frequent awakening.

Figure 4

Neuropsychiatric problems in adulthood.

Other findings

Visual problems were reported in 4 out of 13 individuals (31%). Patient 8 had cortical blindness despite the lack of lesions on brain MRI. Patient 11 had bilateral optic nerve atrophy, and was born prematurely at 34 weeks gestational age. Medical files described a history of neonatal asphyxia and haemorrhage, but cerebral imaging was unavailable. Other described visual problems were severe visual impairment, convergent strabismus, amblyopia and astigmatism. Four individuals had problems with feeding such as difficulties with chewing (patient 10) and dysphagia (patient 6 and 12). One patient had a percutaneous endoscopic gastrostomy tube (patient 11). No cardiac abnormalities were reported.

Three patients (patients 3, 4 and 5) in our study carried the previously described recurrent variant, c.619C>T, p.Arg207Trp.20–23 All seemed generally less neurologically affected in adulthood, with only limited abnormalities on clinical examination. All three individuals talk fluently with normal vocabulary and can walk independently. They all had special education and can work. Patient 5 became seizure-free at the age of 22 years, patient 3 at the age of 40 years and patient 4 still has monthly seizures at the age of 24 years.


The current study analysed the course of illness and current phenotypes of 13 adult patients with KCNQ2 encephalopathy. Because of the small sample size, the high degree of allelic and phenotypical heterogeneity, and our retrospective method, this should be regarded as an early effort to begin understanding the KCNQ2 adult outcome spectrum. Another issue limiting generalisation is that clinical care of infants with KCNQ2 encephalopathy has changed considerably since the years when this cohort was born, and potentially might influence outcome. Despite these limitations, we highlight three broader conclusions about the adult outcomes, if only to provoke more investigation, namely: (1) Seizures become far less frequent, but breakthrough events occur in many patients, (2) Intellectual, behavioural, motor and language disabilities persist despite seizure remission, (3) Outcome severity remains correlated with the pathogenic variant.

Neonatal seizures in individuals with KCNQ2 encephalopathy are highly therapy-resistant at first presentation, and earlier studies6 33 37 showed a high rate of seizure remission within the first year of life. The longer perspective of our study confirms that seizure frequency decreases with age, although the age of clinically relevant epilepsy improvement varied from the first month of life to the age of 8 years (median: 12 months). Importantly, seizure-free periods of 2 years or more followed by recurrence of seizures were seen in over a third of the individuals in our study. Previously described children with seizure freedom in the first year of life might have shown later seizure recurrence in longer follow-up studies. The large majority (10/13; 77%) of adult patients was indeed seizure-free for 2 years or longer at the time of study entry, but most of them (50%) reached prolonged seizure freedom between 11 years and 20 years of age only. Seizure type evolved with age, from focal tonic seizures being most prominent in early infancy, to focal impaired awareness seizures and unknown onset tonic-clonic seizures in older children and adults. Noteworthy, patient 11 was born prematurely at 34 weeks gestational age. She was the only patient still having daily seizures in adulthood. Suspicion of neonatal asphyxia and haemorrhage was reported and could have influenced the outcome. Unfortunately, no information on cerebral imaging was available to confirm the presence of perinatal damage. The patient was however reported to have optic nerve atrophy, a symptom not previously reported in KCNQ2 encephalopathy, and most likely caused by her premature birth and hypoxic-ischaemic encephalopathy.38 39

Evidence for favourable response to SCBs has been described in several previous studies and it is generally accepted that they should be considered as first-line treatment in young patients presenting with seizures and KCNQ2 encephalopathy.15 33 37 40–42 The design and sample size of our study does not allow to draw strong conclusions about treatment response. However, reflecting this clinical consensus, all but one of the seizure-free patients still taking ASM were currently on an SCB. Voltage-gated sodium channels and KCNQ potassium channels co-localise and are bound at critical locations of the neuronal membrane.43 Therefore, the efficacy of SCB could be linked to their modulating effect on this channel signalling complex. Maximising potential therapeutic benefits for KCNQ2 encephalopathy may depend on early genetic diagnosis and effective seizure control at disease onset, and is best tested via prospective controlled trials.

Seizure control notwithstanding, most patients in this series continued to exhibit severe ID as adults. Although moderate and mild ID was observed in a few, all adults needed assistance in daily life activities. Nearly half (46%) of the individuals did not develop any speech or walked independently. Additional neuropsychiatric problems in adulthood were common (figure 4), and 77% of the individuals had either a formal diagnosis of autism spectrum disorder or showed autistic features. The brain mechanisms driving persistent ID, autism and other behavioural disabilities in KCNQ2 encephalopathy are poorly understood. Studies introducing human alleles into model animals may address this gap in knowledge. In one recent study, mice heterozygous for a Kcnq2 null allele, equivalent to a large group of patients with BFNE, showed, in addition to elevated susceptibility to chemoconvulsant-induced seizures, behavioural abnormalities including autism-associated behaviours such as reduced sociability and enhanced repetitive behaviour.44 Behavioural characterisation of mice heterozygous for Kcnq2 Thr274Met, a highly recurrent KCNQ2 encephalopathy allele, revealed deficits in spatial learning and memory in adults but no gross abnormality during early neurosensory development.45 Rigorous parallel investigation of the developmental trajectories in both humans and model animals is needed to uncover the mechanisms.

No other non-neurological abnormalities were reported in our cohort, which is in line with the fact that KCNQ2 is only expressed in the central and peripheral nervous systems.

Three unrelated subjects in our study (patients 3, 4 and 5) carrying the recurrent missense KCNQ2 variant, c.619C>T, p.Arg207Trp, displayed neurodevelopmental problems not compatible with a diagnosis of B(F)NE but had a remarkably milder phenotype compared with other individuals in the cohort. All had received special education and were at the time of enrolment working in a sheltered workshop or as a part-time volunteer. One patient transmitted the variant to a son who also has a phenotype compatible with KCNQ2 encephalopathy. Although significantly disabled, these individuals appear relatively similar to each other, and as a group were more mildly affected than the other study subjects with de novo variants. This suggests that disease outcome depends to some extent on the underlying variant, which is in line with previous studies.7 Remarkably, Arg207Trp has been reported before in BFNE families including individuals with myokymia20 or severe myoclonus-like dyskinesia.21 These features were not seen in the individuals described in our study, although none underwent electromyography. Patients with myokymia experience involuntary rhythmic generalised muscle contractions provoked by hyperexcitability of the lower motor neurons. KCNQ2 is indeed also expressed by spinal motoneurons.20 Functional studies showed that this variant has a unique functional effect on M-current, markedly slowing opening and closing kinetics, which is in some respects intermediate between the mild current suppression exhibited by variants found in BFNE and the strong suppression characteristic of many variants leading to KCNQ2 encephalopathy. We show that some individuals with this variant show a mild to moderate degree of ID. This highlights how KCNQ2-related epilepsies now appear as a disease spectrum rather than a dichotomous entity (BFNE vs KCNQ2 encephalopathy). It also shows that despite this increased complexity, the specific KCNQ2 variant remains a central factor contributing to outcome, further influenced by environmental and/or genetic factors. In this respect, it is worth noting that the first descriptions of patients with KCNQ2 encephalopathy were based on sequencing of genetically unsolved patients that were previously diagnosed with MRI-negative early infantile epileptic encephalopathy (EIEE).6 30 EIEE is defined clinically by burst-suppression EEG, refractory seizures, evolution to hypsarrhythmia, poor global developmental progression and high risk of early mortality.46 Recent case series continue to show KCNQ2 variation as the most frequent genetic cause of EIEE, or Ohtahara syndrome.1 26 The current study of adult patients, along with evidence from other work, however emphasises how this patient subgroup has diverged the classical EIEE syndrome in early features, treatment responsiveness, epilepsy evolution and neurodevelopmental outcome.

Some specific limitations of this study were intrinsic to the approach—data were collected retrospectively from the medical records. Although we clarified issues by contact with current treating physicians, in many instances, older data may be limited in detail or impossible to further validate. We were unable to apply a uniform criterion for the definition of epilepsy improvement since we depended on the impressions and interpretation of the treating physician at prior times, in diverse settings. Similarly, we based seizure classification on the description in the medical file of the patient. Often it was unclear, based on available descriptions or routine EEGs, whether seizures had a focal or generalised onset.

In conclusion, in this study, patients with KCNQ2 encephalopathy were mostly seizure-free in adulthood. Years-long seizure-free periods interrupted by seizure recurrences were frequently seen during childhood and adolescence, but seizure frequency generally declines with age. Most adult patients have a severe ID, and language, motor and behavioural problems are still a concern during adulthood. Systematic, prospective natural history studies on KCNQ2 encephalopathy should ideally extend into adult life but until the current ongoing improvements in early diagnosis and treatment are implemented more uniformly, this will add an additional dimension of heterogeneity that will challenge analyses of aggregate outcomes.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information.

Ethics statements

Patient consent for publication

Ethics approval

We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines. Because this is a retrospective study that only involves reporting of anonymised historical clinical data of patients, our ethical committee (University Hospital Antwerp/University of Antwerp) does not require a formal request of an ethical approval. We did however obtain explicit verbal informed consent for inclusion in the study from all parents or legal guardians of patients.


The authors thank the referring physicians and the participating families for their cooperation and support of this research.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Contributors SW, SB: design of the study, acquisition and interpretation of data, drafting of the manuscript, revision of the manuscript for intellectual content. EC: acquisition and interpretation of data, drafting of the manuscript, revision of the manuscript for intellectual content. KMJ, AD, FM, GR, GL, LV, MKK, PS, RSM: acquisition of data, revision of the manuscript for intellectual content.

  • Funding EC received support for this work from the Jack Pribaz Foundation and the Miles Family fund. PS developed this work within the framework of the DINOGMI Department of Excellence of MIUR 2018–2022 (Law 232 of 2016). SW received support from FWO-FKM (1861419N), Jack Pribaz Foundation, KCNQ2 Cure and KCNQ2 e.v. The funders supported the (co)authors for their time dedicated to this study, but were not involved in study design, collection, analysis and interpretation of data, writing of the manuscript, or the decision to submit the manuscript for publication.

  • Competing interests EC is a consultant to Xenon Pharmaceuticals and Knopp Biosciences; his participation in this work has been reviewed and approved by Baylor College of Medicine in accordance with institutional conflict of interest policies. MKK serves on speakers bureaus for Greenwich, Novartis and Lundbeck, and on an advisory board for Stealth Biotherapeutics. PS received fees from Ultragenyx, Zogenyx, Biomarin, PTC pharmaceuticals, GW pharma, Neuraxpharma and research grants from GW pharma, PTC Pharmaceuticals, ENECTA SV, Kolfarma, and has been investigator for clinical trials for Ultragenyx and Zogenix. SW received speaker and consultancy fees from UCB, Xenon, Zogenix, Lundbeck and Biocodex.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.