Tailored treatments: sensors for personalised medicine

Continous monitoring of medical conditions could help doctors to develop individualised treatments.

Tiny sensors could one day replace bulky ECG equipment to enable continuous vital sign monitoring.

It seems our mothers were right – in our own way, we are all special. Unfortunately when it comes to medical science and devising treatments for everything from diabetes to cancer, this degree of variation between individuals with their differing natural hormone levels, metabolisation rates for drugs, and responses to stress proves problematic.

Doctors have long known that creating a tailored approach to treatment should greatly improve patient outcomes. However, when it comes to diagnosing disease, medics are still dependant on single point-in-time measurements of factors such as blood glucose and hormone levels, which do not show how these factors dip or rise into danger zones over the course of a day. In the same way, monitoring heart rates over time within a hospital does not always show up the physical responses caused by the stresses of everyday life.

Work is therefore taking place on the development of sensors for continuous monitoring, giving medics an ongoing insight into a patient in order to track the spread of disease, monitor their exposure to environmental factors, assess their mental health, and so provide a complete picture of their state of being. This data can then also be compared to the biomarkers within a person’s DNA that suggest how – or whether – their body will respond to particular compounds and metabolise certain drugs.

Eventually, once treatments can be made more personal and so more effective, it is hoped there will be both reduced mortality rates and significant savings for healthcare providers such as the NHS. The Wellcome Trust and EPSRC have long been partnering to support both centres of excellence for medical engineering and schemes such as the Innovative Engineering for Health fund.

‘The main advance (from these) is technology to remotely detect vital signs in patients via a webcam,’ said Dr Meher Antia, business analyst in the technology transfer division of the Wellcome Trust. ‘Oxford University’s Institute of Biomedical Engineering (IBME) has recently spun out a company called Oxehealth to commercialise this.’

Webcam images can be used to calculate heart rate and breathing.

Oxehealth’s system monitors respiratory rate, pulse rate and oxygen saturation by measuring the light reflected from an individual’s face using a webcam. Although the basics of this have been used clinically since the 1980s to monitor hospital patients using a finger probe sensor with in-built LEDs and a photodiode, interference caused by ambient, artificial light has hampered the application of the technology. The Oxford invention includes novel algorithms that remove the effects of ambient light interference, allowing the technology to be used in everyday settings such as a patient’s home. Other algorithms are used to process the image recorded with the webcam and extract heart rate, respiratory rate and oxygen saturation.

Work is also progressing to monitor chemical markers. ‘Continuous monitoring, particularly as part of the management of diabetes, is an important trend,’ said Professor Tony Cass of the Department of Chemistry and Institute of Biomedical Engineering at Imperial College London. ‘When monitoring a condition such as this, doctors must currently rely on data from a small number of measurements taken several times per day. However, many of the factors to be measured change on a much smaller timescale so must be measured continuously in order to get an accurate picture of what is going on.

‘Finding out what might be unusual for a person, rather than unusual compared to a clinical average, is very useful when deciding if they need treatment.’

‘Aside from glucose monitoring, continuous monitoring also has an important role to play in fertility. The two hormones FSH (follicle-stimulating hormone) and LH (luteinizing hormone) produced by the pituitary gland oscillate in order to make a woman fertile. However, it has recently been discovered that FSH may have a 95 minute cycle, and if this is not occurring, it is likely that the person involved will be infertile. Other hormones such as cortisol and melatonin also have a daily rhythm if properly present. Monitoring is usually carried out in a research context, and natural levels differ between individuals. Finding out what might be unusual for a person, rather than unusual compared to a clinical average, is very useful when deciding if they need treatment.’

So attractive has the issue become that it is even the subject of an XPRIZE. Consisting of two separate and consecutive competitions over two years, the Nokia Sensing X Challenge launched in April this year with a goal of aiding development of a wide array of new sensors for the mobile health industry. Cheshire-based ABI-med are involved in the first competition. They have created advanced miniature ECG devices and software, tailored to continuously monitor a patient’s heart rate and perform tasks such as alerting doctors of any abnormalities using any Android phone, laptop or tablet PC via WiFi or USB.

‘The ABI-med model is superior to the existing ECG devices in the sense that it has excellent recording and signal quality, which is not easily affected by the movements of the patient,’ said Dr Constantinos Anagnostopoulos, head of research and development. ‘However, our aim now is to further deliver to the market more and more innovative and miniature products, which will be able to operate without interruption whilst delivering a diagnostic grade ECG and thus drastically improving the quality of monitoring.’

Imperial’s sensor penetrates the skin but doesn’t reach pain receptors.

However, there are still a number of challenges to face before some types of sensor are widely used. ‘There hasn’t been a great deal of success in implanting sensors for home use,’ said Imperial’s Cass. ‘We have to reconcile the benefits with the problems that people experience, in that the sensors don’t work well for very long. The body is a hostile environment and implanting an active device is much more problematic than implanting something like an artificial knee. The act of molecular exchange that takes place during monitoring means the sensors are more likely to be rejected, and this is a big challenge for the whole field.’

However, he believes this can be overcome. ‘Our approach has been to develop a minimally invasive device a few microns below the skin so it doesn’t even reach the capillaries – most of the body’s defence systems are in the blood,’ he explained. ‘This interstitial region also has the benefit of not containing any pain centres, so there is minimal trauma and risk of infection. The patient can insert the sensors, then remove them 24 hours later – it’s like wearing contact lenses.

‘Once they are mass produced, the cost will reduce and we hope they will become the equivalent of the diabetic finger-stick, in terms of price. Putting a different chemistry on the sensor will allow other substances to be monitored, too. We could also look at how therapeutic drugs are metabolised, and tailor doses to suit the individual. Combining this information with a person’s genetic profile will give a much fuller picture than just looking at the genome itself, especially when it comes to clinical trials.’

The Imperial group have scheduled patient trials to begin by the end of 2013. Although development of such sensors and sensing systems faces challenges, new interest and funding opportunities are producing results. It appears this might be the beginning of the end for the one-size-fits-all approach to treatment after all.