Predicting progression in Parkinson’s disease

Background Early prognostic stratification in Parkinson’s disease (PD) remains challenging at the time of diagnosis when only limited clinical information is routinely available. In this study, we aimed to identify, at diagnosis, clinical predictors of progression to Hoehn and Yahr stage 3 and to develop and externally validate a clinically applicable prognostic model based on routinely available clinical features. Methods We analysed data from a single-centre Italian cohort of 301 patients with PD followed prospectively for a median of 7.0 years (IQR, 3.0–10.0) from diagnosis. Time to progression to Hoehn and Yahr stage 3 was assessed using Cox proportional hazards regression. Clinical variables associated with progression were combined in a multivariable Cox model to estimate predicted probabilities of progression to Hoehn and Yahr stage 3 at 3, 5 and 10 years from diagnosis. External validation was performed in an independent cohort of 635 newly diagnosed, drug-naïve patients from the Parkinson’s Progression Markers Initiative. Results Older age, bilateral symmetric parkinsonism and rapid eye movement sleep behaviour disorder (RBD) at diagnosis were associated with faster progression to Hoehn and Yahr stage 3. The prognostic model based on these three clinical variables demonstrated good discriminative performance for predicting progression to Hoehn and Yahr stage 3 in both the development and external validation cohorts. Conclusions Clinical features routinely available at the time of diagnosis can support early risk stratification for motor progression in PD. Such information may support prognostic counselling and inform clinical decision-making from the first neurological evaluation.