Prognosis in Parkinson’s disease: an individual-patient-data meta-analysis of six European incidence cohorts

Background: An accurate understanding of prognosis in Parkinson’s disease (PD) is important for patient information provision, personalised treatment, and clinical trial design, but most previous research has been biased towards younger, healthier patients. Objectives: To describe key clinical outcomes longitudinally and identify baseline prognostic factors (predictors) for these outcomes using population-representative PD cohorts. Methods: We meta-analysed individual-patient-data from six incidence cohorts in Western Europe (Norway, Sweden, and UK). Each cohort aimed to recruit and follow-up all newly-diagnosed cases in defined population/incidence periods (between 2000–2011). We described postural instability (Hoehn & Yahr stage 3), functional dependency (needing help with daily activities), dementia, and death with up to 12 years’ follow-up and investigated clinical and genetic predictors using frailty Cox models. Results In 883 population-based incident patients, median time to postural instability and functional dependency was 7.4 years. Dementia affected 49.6% by 10 years and 54.7% had died by 12 years (median survival 9.4 years). Older age, higher MDS-UPDRS part III score, and lower MMSE were significantly associated with all outcomes; cognitive symptoms and GBA polymorphisms with each outcome except mortality; and APOE ε4 with increased mortality and dementia. Conclusions: This first individual-patient-data meta-analysis of population-based incidence cohorts provides robust prognostic data, with fewer selection biases than previous PD studies, for informing people with PD about prognosis. In incidence cohorts, overall PD prognosis in worse than previously suggested, with key outcomes often occurring early. Further work should develop validated prognostic models for objective stratification of prognostic risk and for personalised medicine.