A team of scientists from University College London (UCL) and University Medical Center Göttingen has developed a blood test using artificial intelligence (AI) to predict Parkinson’s disease up to seven years before symptoms appear. Parkinson’s, which affects around 10 million people worldwide, is the fastest-growing neurological disorder globally.
The disease progresses as nerve cells in the brain’s substantia nigra, responsible for movement, die off due to the accumulation of the protein alpha-synuclein. Currently, patients are treated with dopamine replacement therapy only after symptoms such as tremors, slowed movement, and memory issues manifest. Early prediction and diagnosis are crucial for developing treatments to protect dopamine-producing brain cells.
Senior author Professor Kevin Mills (UCL Great Ormond Street Institute of Child Health) emphasized the importance of diagnosing patients before symptom onset. “We need to protect our existing brain cells. As new therapies emerge, early diagnosis is vital to begin treatment before symptoms appear,” Mills said.
Published in ‘Nature Communications’, the research employed machine learning to analyze a panel of eight blood-based biomarkers, achieving 100% accuracy in diagnosing Parkinson’s. The team then tested the AI’s predictive ability on 72 patients with Rapid Eye Movement Behavior Disorder (iRBD), a condition where patients physically act out their dreams and are at high risk of developing Parkinson’s. The AI identified that 79% of iRBD patients had a profile similar to Parkinson’s patients.
Over a ten-year follow-up, the AI predictions matched the clinical conversion rate, correctly identifying 16 patients who developed Parkinson’s up to seven years before any symptoms appeared. The researchers continue to monitor these patients to further validate the test’s accuracy.
Co-first author Dr. Michael Bartl (University Medical Center Göttingen and Paracelsus-Elena-Klinik Kassel) highlighted the test’s potential to allow earlier drug interventions, potentially slowing or preventing disease progression. “By identifying eight proteins in the blood, we can detect potential Parkinson’s patients years in advance,” Bartl said.
Co-author Professor Kailash Bhatia (UCL Queen Square Institute of Neurology and National Hospital for Neurology & Neurosurgery) and his team are currently testing the blood test’s accuracy on high-risk populations, such as those with genetic mutations linked to Parkinson’s.
Professor David Dexter, Director of Research at Parkinson’s UK, praised the research, co-funded by Parkinson’s UK, as a significant advancement. “This blood test is much less invasive than lumbar punctures, which are increasingly used in clinical research,” Dexter highlighted. He added that the test could potentially differentiate Parkinson’s from other conditions with similar early symptoms, such as Multiple System Atrophy or Dementia with Lewy Bodies.
The findings contribute to the ongoing search for a simple, definitive diagnostic test for Parkinson’s disease, marking a major step forward in early detection and treatment.