AI-enhanced blood test may detect Parkinson’s years before onset

<span>An ageing global population means Parkinson's disease is the fastest growing neurodegenerative condition in the world.</span><span>Photograph: Joe Giddens/PA</span>
An ageing global population means Parkinson's disease is the fastest growing neurodegenerative condition in the world.Photograph: Joe Giddens/PA

A blood test that draws on artificial intelligence can predict who will develop Parkinson’s disease up to seven years before symptoms arise, researchers say.

The test is designed to work on equipment already found in many NHS laboratories and, if validated in a broad population of people, could be made available to the health service within two years.

There are no drugs to protect the brain from Parkinson’s at present, but an accurate predictive test would enable clinics to identify people who stand to benefit most from clinical trials of treatments that aim to slow or halt the disease.

“At the moment, we’re shutting the stable door after the horse has bolted,” said Prof Kevin Mills, a senior author on the study at UCL Great Ormond Street Institute of Child Health. “We need to get to people before they develop symptoms. It’s always better to do prevention rather than cure.”

Parkinson’s disease is the world’s fastest growing neurodegenerative condition, a trend driven primarily by the ageing population. The disorder affects more than 150,000 people in the UK and 10 million worldwide. It is caused by the buildup of a protein called alpha-synuclein that damages or destroys nerve cells which produce an important substance called dopamine in part of the brain called the substantia nigra.

People who develop Parkinson’s can experience tremors, difficulties with movement and muscle stiffness, but also problems with balance, memory, dizziness and nerve pain. Many receive dopamine replacement therapy, but efforts are under way to find treatments that slow or stop the disease.

To develop the test, scientists at UCL and the University of Göttingen used a machine learning algorithm to spot a signature pattern of eight blood proteins in patients with Parkinson’s. The algorithm was then able to predict future Parkinson’s in other patients who provided blood samples. In one patient, the disorder was correctly predicted more than seven years before symptoms arose. “It is possible that it could go back even further,” said Dr Jenny Hällqvist, at the UCL Institute of Neurology, and first author on the study published in Nature Communications.

Prof Roger Barker, a consultant neurologist who specialises in Parkinson’s at the University of Cambridge and Addenbrooke’s hospital, said if validated by other groups, the test raised the possibility of diagnosing Parkinson’s at the very earliest stages, enabling patients to be enrolled in clinical trials when the disease process had just begun. “As such, we could treat people with Parkinson’s with disease-modifying therapies before they have lost many cells in their brain,” he said. “Obviously, we still need to find such therapies, but this study is a step in the right direction.”

Related: ‘In some ways, Parkinson’s has meant a new lease of life’: meet the Movers and Shakers

Prof Ray Chaudhuri, the medical director of the Parkinson Foundation International Centre of Excellence, said there was a “massive unmet need” for blood tests that predict and diagnose Parkinson’s, but cautioned that such tests come with “major challenges”.

“Parkinson’s is not a single disease but a syndrome and can present in various different ways,” he said. “As such, management differs and one size does not fit all. The prediction is unlikely to signpost these subgroups at this stage.” Without effective treatments an early diagnosis raises considerable ethical issues, he added, as well as potentially affecting patients’ insurance policies.

“The process does help us have a group of people with Parkinson’s who may be ready or suitable for future trials of neuroprotective molecules,” Chaudhuri said. “Furthermore, there is some preliminary evidence that in such “at risk” people with Parkinson’s, physical activity and programmed exercise may be beneficial in terms of potentially slowing the course of the illness.”