CSF aß42 is related to cortical metabolism (FDG-PET) in non-demented parkinson’s disease patients

CSF Aß42 is an established biomarker for Alzheimer's disease (AD), but little is known about a possible role of amyloid dysmetabolism in PD CNS function. However, amyloid cross pathology is frequently seen in patients with PD dementia, and PD patients may have slightly reduced CSF levels of Aß42 as compared to controls. Herein, we wanted to examine the relationship between the two established AD biomarkers CSF Aß42 and FDG-PET in a population of early PD. Patients with early Parkinson's disease (PD) were recruited. Subjects had to fulfill the Gelb clinical diagnostic criteria for probable PD, DaTSCAN had to be pathological, Hoehn and Yahr stage < 3 and disease duration ≤ 6 years. Exclusion criteria were dementia and other somatic or psychiatric diseases that might contribute to cognitive impairment. 30 subjects successfully underwent lumbar puncture, 3T magnetic resonance structural (MRI) and FDG-PET. CSF Aß42 was analyzed with ELISA (Fujirebio Europe, previously Innogenetics) in all patients. FDG-PET was normalized against the cerebellum. Structural MRI and FDG-PET was processed and analysed using FreeSurfer. Surface based statistics were performed, correlating the imaging parameters against CSF Aß42 while adjusting for age, gender and multiple comparisons. Mean age was 64.9 years of age (range 47-74). Mean disease duration was 2.5 years. Mean UPDRS part III motor score was 14.9. Thirteen subjects had normal cognition, four had subjective cognitive complaints and fourteen had MCI (as determined by neuropsychiatric examination). Mean CSF Aß42 was 943 (range 707-1280). All PD patients had CSF Aß42 levels above diagnostic cut-off for Alzheimer's disease (600 ng/mL). Corrected for age, gender and multiple comparisons, the statistical map showed widespread correlations between cortical metabolism and CSF Aß42. There were no correlations between cortical thickness and CSF Aß42. CSF Aß42 levels correlate with reduced cortical metabolism in a population of early PD patients without dementia. Putatively, amyloid dysmetabolism may be related also to PD cognitive impairment, though CSF Aß42 levels were above cut-off.