What can a smart watch do to help those living with Huntington’s disease? And what might it mean for patients diagnosed with Alzheimer’s, Multiple Sclerosis and other debilitating diseases? Mark Greener finds out
Huntington’s disease (HD) – a devastating, disabling and deadly neurodegenerative condition – characteristically causes a jerky, involuntary, dance-like ‘chorea’ of the arms and legs. Chorea, along with HD’s diverse other symptoms, dramatically undermines a patient’s ability to carry out normal daily activities that the rest of us take for granted. Currently, however, the routine assessment of a HD patient offers – at best – only a subjective snapshot of their wellbeing and functional ability, which hinders clinical assessments and complicates drug development.
However, Teva Pharmaceutical Industries is collaborating with Intel to develop a wearable-tech platform based on a sensor-equipped smartwatch, smartphone and a unique data analytics engine that continually monitors and analyses HD’s hallmark motor symptoms. For Spyros Papapetropoulos, Global Development Head, Neurodegenerative Diseases and Movement Disorders at Teva, the imminent start of clinical testing represents the culmination of 13 years’ work.
While HD might seem to be a niche market, the combination of wearable technology and artificial intelligence (AI) could find numerous other clinical applications as well as helping to speed up the development of much-needed medications for several neurodegenerative diseases. ‘HD is a platform that allows us to assess the value and features of our unique technology,’ says Dr Papapetropoulos. ‘We anticipate that the technology will have much wider applications than HD alone.’
An abnormal gene on the short arm of chromosome 4 causes HD. The healthy gene contains a run of about 20 repetitions of three nucleic acids: cytosine (C) followed by adenine (A) and then guanine (G). The abnormal gene contains extra CAG repeats: HD begins to emerge when the gene contains about 35 to 40 CAG repeats. Because the gene is dominant, sooner or later everyone who lives long enough with the abnormal gene develops HD. In addition, the children of a person with HD have a 50:50 chance of inheriting the gene.
According to a recent review, HD prevalence varies more than 10 fold across the world. Rates are highest in North American (7.33 per 100,000) and the UK (6.68 per 100,000) and lowest in Asia (0.40 per 100,000). Nevertheless, HD seems to be becoming more common – in some parts of the world at least. HD prevalence increased by 20.1 per cent and 15.5 per cent per decade in North America and the UK respectively, for example.
Despite being relatively rare, HD’s course is well defined. The average age of HD onset is about 40 years. In addition to chorea and other movement disorders, HD patients develop more subtle problems, including psychiatric symptoms – such as psychosis, depression and obsessive compulsive disorder – and progressive cognitive impairment. Although the speed of the decline varies, a person with HD inexorably descends into disability and eventually death. Indeed, HD usually proves fatal 10 to 15 years after diagnosis. This predictable clinical profile inspired Teva to assess the value of wearable technology combined with AI in monitoring the disease.
‘We know what to expect in terms of clinical progression,’ Dr Papapetropoulos says. ‘While HD patients experience a wide range of symptoms, including behavioural and cognitive complications, chorea and other motor symptoms have a great impact on their lives and can be objectively measured. This made HD the perfect condition to explore and evaluate the use of wearable technology that measures abnormal movement.’
The wearable, watch-like device uses accelerometers – sensors that measure tilting motion and orientation – to continually monitor HD’s motor symptoms. A linked smartphone (patients don’t need to carry the phone around with them) uploads the data to Intel’s Cloud-based ‘machine learning platform’ – part of their open-source Trusted Analytics Platform of Big Data applications. The platform then translates data from each patient into objective scores and visual representations of motor function in the wrist and trunk. The rich data collected by the wearable device can improve clinicians’ understanding of the disease’s progression and impact, as well as facilitating the evaluation of new treatments.
‘The information collected could, when validated, act as “digital biomarker” in clinical studies. This should allow us to progress a drug to market more rapidly,’ Dr Papapetropoulos says. ‘In addition, the information could improve HD care. Typically, a HD specialist sees a patient once every six months. You have to cram the experience of 4,400 hours of living with HD into a 30-minute consultation. The hospital setting is very artificial, and some people may feel anxious or seek to impress the doctor, both of which can influence motor function. Moreover, the current assessment tools are subjective, and doctors can inadvertently be biased. The wearable device could provide much more accurate and objective information that will help the consultation, aid treatment decisions and hopefully reduce pressure on healthcare resources. I would expect that, depending on the success in the studies, the device could reach the clinic in between three to 10 years.’
Teva is clearly confident that the technology will work. It has jumped the usual pilots, and the device will begin clinical testing in a sub-study of 60 patients enrolled in the Open-Pride HD, an open-label extension phase II trial, which is due to begin later this year in the USA and Canada.
Open-Pride HD will assess the long-term effects of pridopidine – an investigational drug being developed for the treatment of HD. The study follows the recently completed Pride-HD study; a 52-week, dose-ranging trial of pridopidine twice daily versus a placebo. Pridopidine is an agonist of sigma-1 receptors (S1R), which influence neuronal function by increasing production of brain-derived neurotrophic factor (BDNF). In turn, BDNF maintains healthy neurones and supports neurogenesis (the growth and development of nervous tissue). BDNF levels are decreased in several neurodegenerative disorders including HD, Parkinson’s disease, Alzheimer’s disease and amyotrophic lateral sclerosis.
‘The first step is to ensure that patients can use the technology and validate the data. The Open-Pride HD study will allow us to examine the device in a very structured, robust way,’ Dr Papapetropoulos says. For example, the study will collect data using the wearable sensor during standard neurological assessments, such as hand rotation, which measures motor speed and control, and the timed ‘up-and-go’ test (where the patient rises from a chair, walks a set distance, such as three metres, and then returns to the chair and sits down). This should provide the data needed to train the algorithms and validate the device. Patients will also keep a medication diary that allows Dr Papapetropoulos’ team to examine any effect of treatment on movement measured by the device. ‘All being well, we should get the first results within a couple of years,’ he says.
For Dr Papapetropoulos, the start of testing marks 13 years of hard work. He has been a member of several research networks across industry, academia and non-governmental organisations that agreed to develop methods that could objectively measure motor function in the daily life of patients with neurological disorders for several years. However, until recently, the computing power and AI algorithms needed to analyse the terabytes of data were not available.
An accelerometer in the wrist device, for example, samples movement at 50Hz – essentially 50 data points a second – and the study is predicted to last six months. That’s about 800 million data points per sensor. However, as Dr Papapetropoulos points out, the smart phone contains accelerometers capable of measuring movement at even higher frequencies. In other words, there’s considerable potential to enhance the resolution of wearable technology in the future.
Dr Papapetropoulos predicts that the combination of wearable device and AI algorithms will find numerous applications other than HD. ‘The technology could also be used to collect information about slowness of movement, tremor and stiffness in Parkinson’s disease,’ he says. ‘Indeed, many machine learning algorithms used in Huntington’s disease come from Parkinson’s research.’
Teva is also looking to combine companion digital applications with pharmaceuticals across their portfolio ‘wherever relevant’. For instance, it is introducing digitally enabled inhalers for asthma and chronic obstructive pulmonary disease that remind patients when a dose is due, track use and upload the data to smartphones. So, the inhalers improve adherence and facilitate data sharing between patients and their healthcare providers. Other Teva projects are beginning to explore digitally enabled solutions in areas such as migraine, MS and psychiatric conditions.
‘Teva is looking to develop smart solutions for every therapeutic area we work on,’ Dr Papapetropoulos concludes. ‘The value propositions are still evolving. But it’s clear that the combination of wearable technology and machine learning platforms can complement future trials in several diseases. More importantly, the combination offers an unprecedented opportunity to provide meaningful and actionable feedback to doctors, patients and caregivers. It’s an exciting, empowering prospect.’
You're the expert! Write for The Engine or share your articles, papers and researchAdd your content
Add your content
Sign up for Ignition, our regular, ideas-packed newsletter