Cerebrospinal fluid concentrations of inflammatory markers in Parkinson's disease and atypical parkinsonian disorders

Abstract

Inflammation has been implicated in the pathogenesis of Parkinson's disease (PD). We here investigate levels of inflammatory biomarkers in cerebrospinal fluid (CSF) in PD and atypical parkinsonian disorders (APD) compared with neurologically healthy controls. We included 131 patients with PD and 27 PD with dementia (PDD), 24 with multiple system atrophy (MSA), 14 with progressive supranuclear palsy (PSP) and 50 controls, all part of the Swedish BioFINDER study. CSF was analyzed for CRP, SAA, IL-6, IL-8, YKL-40 and MCP-1 (CCL2) as well as α-synuclein (α-syn), tau, tau phosphorylated at Thr181 (P-tau), Aβ₄₂ and NfL. In this exploratory study, we found higher levels of the inflammatory biomarker SAA in PDD and MSA compared with controls and PD and higher levels of CRP in PDD and MSA compared with PD. YKL-40 was lower in PD compared with controls. There were multiple positive correlations between the inflammatory markers, α-syn and markers of neuroaxonal injury (NfL and tau). In PD, higher levels of inflammatory biomarkers correlated with worse motor function and cognitive impairment. Thus, inflammatory biomarkers were increased in PDD and MSA. Furthermore, inflammatory biomarkers correlated with more severe disease regarding motor symptoms and cognitive impairment in PD, indicating an association between inflammation and more aggressive disease course. However, the results need confirmation in follow-up studies.

Figure 1

Inflammatory CSF biomarkers in different diagnostic groups. Box plots with scatter with levels of inflammatory CSF biomarkers (a) CRP, (b) SAA, (c) IL-6, (d) IL-8, (e) YKL-40 and (f) MCP-1 in the different diagnostic groups presented as median and inter quartile range. Outer whiskers are 1.5 IQR. P values are from univariate general linear model adjusting for age, gender, disease duration, inflammatory condition and total somatic illness. In figure a, there were 4 outliers in PD and 2 in MSA outside the axis limit. In figure b, there were 5 outliers in PD, 3 in PDD, 2 in PSP and 2 in MSA outside the axis limit. In figure c there was 1 outlier in PDD outside the axis limit. Outliers outside the graph limits are included in all statistical analysis.

Figure 2

Correlations between CSF CRP and clinical test scores in PD. Correlations of ln-transformed CSF levels of CRP with UPDRS-3 (a) and CRP with FACIT-f (b) in the PD group. Lines represent linear regression and 95% CI.

Figure 3

Correlations between CSF inflammation biomarkers and clinical test scores in MSA. Correlations of ln-transformed CSF levels of CRP (a) and IL-8 (b) with UMSARS total score in MSA. Lines represent linear regression and 95% CI.