more_vert
SPEECH is an evolutionarily fascinating behaviour that is unique to humans. Ideally, kids develop speech seamlessly simply by exposure to their native language. Yet an astounding half a million Australian children are seeking clinical care for speech and language disorders at any one time (here and here). This far outnumbers the more commonly recognised neurodevelopmental disorders such as autism (164 000 in 2015).
Speech and language outcomes are typically first identified between 2 and 3 years of age, based on common “red flags” of symptomatology or behaviour.
Families can spend hours every week in speech therapy, sometimes for months or even years of a child’s life, working on surface symptoms and creating a drain on services, with no guarantee of a good outcome.
Concerned parents visit their GP or paediatrician, and are referred on to speech pathology or other early intervention services. One-third of referred children will resolve their communication challenges without further long term impact (and arguably may not have required services in order to resolve). One-third will mainly resolve their problems but will require extra support at core times of transition or learning demands. The remaining third will go on to have a chronic speech or language disorder with associated adverse outcomes.
How can we more precisely predict which children will respond to therapy and which will resolve by themselves?
At present, an inappropriate “watch and wait” surveillance approach is taken for some children who would have benefited from earlier intervention. Conversely, many children are unnecessarily accessing already limited services.
To address this problem, a large cohort study of language at the Murdoch Children’s Research Institute produced the most robust predictive model by collecting and analysing a rich symptom-based and environmental dataset. However, even this gold-standard resource only explained between 30% and 58% of the variance in child language outcomes at ages 4 and 7 years respectively.
Therefore, the missing piece of the puzzle seems to biology – the genetic factors that significantly contribute to speech and language development. It is perhaps unsurprising then that current symptom-based prediction tools and symptom-focused therapies are limited. How can we help clinicians leverage the power of genomics to optimise patient outcomes in debilitating and lifelong speech and language disorders, such as apraxia?
The National Health and Medical Research Council (NHMRC) funded our Australian-led Centre of Research Excellence (CRE) in Speech and Language Neurobiology to begin identifying genetic causes for severe childhood speech disorders. The multinational CRE includes partners at the Max Planck Institute for Psycholinguistics (Netherlands) and the UCL Institute for Child health (UK).
Apraxia is as common as congenital hearing loss, affecting over one in 1000 individuals. However, apraxia has not received anywhere near the equivalent attention or funding because it takes time to appear, it is hard to diagnose and the underlying cause remains elusive.
FOXP2, identified in 2001, was the first gene for speech. Mutations or variants in FOXP2 were associated with a severe apraxia (highly unintelligible speech) but no intellectual disability. Unfortunately, little additional progress has since been made, until now.
Our CRE has recently driven discovery of a plethora of new apraxia-associated genes (Eising and colleagues, Khan and colleagues, Johnson and colleagues).
In line with recent genetic findings in other neurodevelopmental disorders, such as autism and epilepsy, we see significant genetic variation in apraxia. Despite this variation, apraxia genes seem to converge on brain development gene expression pathways, particularly in brain regions underlying speech production.
These discoveries have spurred us to develop preclinical models for therapies targeting underlying causative pathways. With collaborators, we successfully treated a “speech” disorder in a mouse model. We used two distinct drugs approved by the Food and Drug Administration to rescue an ultrasonic vocalisation phenotype in mice lacking the UPF2 gene.
This is the first work in the speech pathology field demonstrating proof of concept of precision therapy in speech pathology. Other preclinical models are in development for further genes associated with apraxia and also stuttering.
For over 400 years, since some of the earliest descriptors of pathological speech phenotypes (Hieronymus Mercurialis, Treatises on the diseases of children, Venice, 1583), we have made little progress in understanding the underlying causes. We have much ground to make up, but recent genetic developments provide significant hope for rapid advances.
We want to drive change to current practice in the longer term by applying genetic risk scores to speech and language prediction tools for GPs and paediatricians that encompass genetic, environmental and symptom-based data points. Our CRE is performing, or contributing to, large genetic association studies to better understand genetic contributions to common speech and language disorders, such as stuttering.
In the short term, we are translating our findings into genomic guidelines to educate paediatricians, geneticists, neurologists and speech pathologists on relevant phenotypes, encouraging referral for genetic testing for the most severely affected individuals.
In turn, this will bring families’ diagnostic journeys to an earlier close, saving time and money on expensive diagnostic tests, and facilitate earlier, more targeted intervention, enhancing outcomes in these children with far-reaching social and economic benefits to Australian society.
Speech and language disorders are debilitating enough, but they also triple the likelihood of poor reading, spelling and maths outcomes and lead to higher rates of school non-completion and restricted employment opportunities (here, here and here). They are also associated with mental health problems and antisocial behaviour; over half of male offenders in the youth justice system have a communication disorder. Further, almost 70% of people with communication disorders are unemployed or in the lowest income brackets.
The huge burden of speech pathology treatment in Australia is conservatively estimated at half a billion dollars annually. Yet despite this outlay, an astounding 50% of children with communication disorders have unmet needs.
Research shows that next generation sequencing produces a diagnostic hit rate of one in three children. Definitive diagnoses allow us to reduce the health care budget, begin looking for treatments, and provide a massive opportunity to optimise outcomes and reduce disability for children with a huge potential to contribute to Australian society.
Professor Angela Morgan is an NHMRC Practitioner Fellow and leads the Speech and Language group at the Murdoch Children’s Research Institute. Professor Morgan’s primary research focus is on speech and language phenotyping in genetic conditions.
The statements or opinions expressed in this article reflect the views of the authors and do not represent the official policy of the AMA, the MJA or InSight+ unless so stated.