Premotor phase biomarkers may enable early PD diagnosis

Premotor phase biomarkers can be as accurate as motor phase biomarkers in diagnosing Parkinson’s disease (PD), offering an earlier diagnosis compared with clinical methods, suggests a study. 

Current clinical diagnostic approaches make it difficult to diagnose PD early, according to the authors, who then sought to determine whether biomarkers processed by machine learning models could allow earlier PD detection. 

In the study, the authors performed a literature search with a 10-year limit, which yielded 161 biomarkers derived from serum, cerebrospinal fluid, and genes. They initially classified these biomarkers into four groups according to the biological timeline of PD pathogenesis and subsequentlyorganized into just two: premotor phase and motor phase biomarkers. 

Finally, the authors carried out two analyses based on these classifications, with the performance of biomarkers assessed using the area under the curve values as derived from machine learning models. 

There were no significant differences seen in either analysis, indicating that all biomarkers can serve as valid diagnostic predictors regardless of their role in the biological sequences underpinning PD pathogenesis. Furthermore, 26 top-performing biomarkers with high area under the curve values (>0.8) were identified. 

“Therefore, our foremost suggestion is further research into the clinical viability of premotor phase biomarkers that compose part of the 26 top-performing biomarkers,” the authors said.