Exploring Research Towards Alzheimer's Disease In CSF Biomarkers Profile

Objective:As is reported that over 50 million people have been diagnosed dementia.China has the largest population,possessing over 18 million dementia patients.AD is the major etiology in dementia,which composes more than half of it.Accurate diagnosis with optimistic intervention could apparently delay the progression and improve the life quality.However,due to significant variability and short of diagnostic modality in clinic practice,accurate diagnosis proves to be a challenge for physicians.CSF biomarker has shown it promising utility here but still need more evidence to support its clinical efficacy.This research intends to conduct CSF biomarkers analysis based on PUMCH dementia cohort,to verify the diagnostic efficacy,to estimate cut-off value,to standardize the sampling,analyzing and storing protocols.In addition,the relationships between biomarkers and clinical characteristics are expected to explore,to provide information in disease process,severity,and prognosis.Methods:Upon PUMCH dementia cohort,recruit individuals consecutively with both dementia and cognitive normal controls.All individuals are assessed with comprehensive medical history,neuropsychological,imaging study.NIA-AA guideline for dementia due to AD is applied to select AD from other type of dementia into AD dementia group,the rest are assigned into non-AD dementia group.CSF samples from pre-scheduled lumbar puncture procedure are collected for ELISA analyses in A ?42,t-tau,p-tau181(Fujirebio,Ghent,Belgium)and NfL(UmanDisgnostics,Japan).The data is further calculated to show the diagnostic efficacy between AD and NAD group by AUC of ROC curve,with relevant cut-off value.Clinic characters,including age,sex,and cognitive level,progressive speed are examined the correlations with CSF biomarkers.Logistic regression is introduced to combine the biomarkers as a whole panel to differential AD dementia.As for quality control,a fixed duration is set to randomly reexamine a small group of samples to disclose the consistence between analyses and stability over long duration of storage.Results:237 individuals recruited from PUMCH cohort and 216 included as 109(50.7%)in AD group,81(37.2%)in NAD group and 26(12.1%)in control group.CSF samples undergo ELISA analysis for A ? 42,total tau,181-P-tau,and 125 for additional NfL test.Diagnostic efficacy evaluation shows that AUCs of ROC between AD and NAD group are 0.68,0.74,0.76,0.63,0.77 and 0.81 in A ?42,total tau,P-tau181,NfL,t-tau/A? 42(TAR)and p-tau181/A? 42(PAR),respectively.Combination utility through logistic regression does not display any clinical improvement with AUC 0.76.In correlation research,age and sex generally not correlate with biomarkers except TAR with sex in all-types dementia.CSF biomarkers prove statistically significant correlation in all-types dementia analysis,and similar in AD group(except A? 42).Cognitive decline rate of follow-up AD patients(n=17)proves independent with initial biomarkers level.As for quality control assessment,reexamination from different batches of kits or after fixed interval displays acceptable stability and consistence with no statistically significant variability.Conclusion:A ?42,t-tau,p-tau 181 all show diagnostic utility,while the ratio t-tau/A?42 and p-tau/A?42 prove higher diagnostic efficacy.Prediction form logistic regression promotes efficacy to 0.76.Biomarkers other than A?42 reveals correlation with cognitive level of AD patients,but prove no prediction power in cognitive decline rate.Although it shows that biomarkers correlate well with cognitive level in dementia patients,we believe it as modifiable statistical bias.ADL seems more reliable than MMSE and MoCA in follow-up.And quality control verify the pre-analytic protocols we adapt can provide generally accessible stability and consistence of data for long-term sample storage and inter-laboratory analysis.