New Tool May Help Researchers Predict Cognitive Decline with Parkinson’s Disease

Predictive Modelling for Parkinson'sA new tool from researchers at Harvard Medical School may allow investigators to predict cognitive decline in people with Parkinson’s disease (PD), which could help design clinical trials and test new therapies. The tool uses an algorithm combining varied personal details — age at onset, baseline score on the Mini Mental State Examination, years of education, motor exam score, sex, depression, and GBA mutation status.

The scientific team, led by Clemens Scherzer, MD, published this work in the scientific journal The LancetNeurology. They compared data from six studies to develop the tool, which they validated with data from three studies, including The Michael J. Fox Foundation’s landmark biomarkers study, the Parkinson’s Progression Markers Initiative (PPMI). Overall, they looked at data from 3,200 people with PD.

Predictive modeling – a process that uses data mining and probability to forecast outcomes – is important as we work toward objective, biological markers of Parkinson’s disease. (In fact, a recent PPMI study did reveal potential biomarkers of Parkinson’s cognitive decline).

Prediction tools can allow studies to enroll fewer participants, thereby saving time and money. If we want to better understand cognitive decline or test a treatment against this symptom, we need a certain number of people who will get cognitive impairment. Without a prediction tool, we may have to enroll a larger population to account for those who won’t develop cognitive decline. The algorithm can help researchers choose people likely to experience this symptom.

An interesting inclusion in the algorithm is education years. The more years of formal education patients in the study had, the greater was their protection against cognitive decline.

“This fits with the theory that education might provide your brain with a ‘cognitive reserve,’ which is the capacity to potentially compensate for some of the disease-related effects,” said Dr. Scherzer. “I hope researchers will take a closer look at this. It would be amazing, if this simple observation could be turned into a useful therapeutic intervention.”

The researchers say the tool is not ready for widespread clinical use as there is considerable work required to optimize the algorithm (their next step) and currently there are no therapies approved to prevent or stop cognitive decline in PD.

Through use in research studies, though, their prediction tool may aid in the discovery of new PD treatments and determine which patients would benefit most from those therapies.

“Prediction is the first step,” said Dr. Scherzer. “Prevention is the ultimate goal, preventing a dismal prognosis from ever happening.”

Read more about predictive modeling of Parkinson’s progression using PPMI data.