Combining biomarkers for prognostic modelling of Parkinson's disease

Background Patients with Parkinson's disease (PD) have variable rates of progression. More accurate prediction of progression could improve selection for clinical trials. Although some variance in clinical progression can be predicted by age at onset and phenotype, we hypothesise that this can be further improved by blood biomarkers. Objective To determine if blood biomarkers (serum neurofilament light (NfL) and genetic status (glucocerebrosidase, GBA and apolipoprotein E (APOE))) are useful in addition to clinical measures for prognostic modelling in PD. Methods We evaluated the relationship between serum NfL and baseline and longitudinal clinical measures as well as patients' genetic (GBA and APOE) status. We classified patients as having a favourable or an unfavourable outcome based on a previously validated model, and explored how blood biomarkers compared with clinical variables in distinguishing prognostic phenotypes . Results 291 patients were assessed in this study. Baseline serum NfL was associated with baseline cognitive status. Nfl predicted a shorter time to dementia, postural instability and death (dementia-HR 2.64; postural instability-HR 1.32; mortality-HR 1.89) whereas APOEe4 status was associated with progression to dementia (dementia-HR 3.12, 95% CI 1.63 to 6.00). NfL levels and genetic variables predicted unfavourable progression to a similar extent as clinical predictors. The combination of clinical, NfL and genetic data produced a stronger prediction of unfavourable outcomes compared with age and gender (area under the curve: 0.74-age/gender vs 0.84-ALL p=0.0103). Conclusions Clinical trials of disease-modifying therapies might usefully stratify patients using clinical, genetic and NfL status at the time of recruitment.

Automated summary

This study aimed to determine if blood biomarkers (NfL and genetic status) can serve as additional clinical measures for prognosis in PD patients. The results showed that NfL and genetic variables have similar predictive ability for unfavorable progression as clinical predictors. The combination of clinical, NfL, and genetic data produced a stronger prediction of unfavorable outcomes compared to age and gender.