论文部分内容阅读
Background: The aim of this study was to establish a causal disease progression model quantifying the sequential change of biomarkers and their relative values as predictors of different Alzheimers Disease stages.Methods: Data used in preparation of this article were obtained from the ADNI database (adni.loni.ucla.edu).Dataset preparation, exploration and visualization were performed using RStudio.Disease progression model were established using extended least squares regression by NONMEM version Ⅶ (Icon Development Solutions, Ellicott City, Maryland, USA).The model building strategy was based on the approach that was widely used in pharmacometrics communities.Results: A total of 398 subjects that had baseline and longitudinal measures of Aβ42, p-tau, hippocampus volume and ADAS-cog were included for modeling analysis.We have successfully established a causal model that has the following characteristics: a) CSFAβ42, p-tau, hippocampus volume and clinical score are all utilized to develop a time line of causal disease progression; b) model can be applied to a large scale of Alzheimer disease spectrum from normal aging to dementia; c) All the potential covariates are tested including demographic characteristics and genotype.MCI is the most widely used indicator to predict progression to AD at present, but having low diagnostic accuracy.Our analysis confirmed this.Results of our study also suggest that Aβ42 or p-tau in isolation is insufficient to explain disease state conversion.However, the ratio of Aβ42 and p-tau, a new indicator we proposed, can be a good predictor.The mathematical expression of the base model is as below (equations1-5):dRatio/dt =K * Ratio * 1-Ratio/Rati max (1)Ration(0)=Aβ0/ p-tau0 (2) Volume =Volume0 + Emax *ratio/ratio+EC50 (3)ADAS-cog =ADAS-cog0-Imax *Volume/Volume+IC50 (4) For AD and LMCIADAS-cog =ADAS-cog0-Imax1 *Volume/ Volume+IC501 (5) For NL and EMCI Conclusions: In summary, the developed causal model was suitable for describing the progression of AD.Covariates that were found to influence ratio baseline were baseline disease state and APOE4 genotype; Factor that influence hippocampus volume baseline was age; factors that influence ADAScog score was baseline disease state.The model could represent a suitable tool for clinical trial simulations and could aid in the design of efficient clinical trials in the future.When ratio is in the range of 7.34 and 8.84, there is a strong possibility that the individual is in a normal condition.When ratio is in the range of 6.26 and 7.34, there is a strong possibility that the individual is in EMCI condition.We should remind patients attention should be paid seriously in lifestyle or mental exercise.When ratio is below 6.26, there is a strong possibility that the individual is in LMCI or AD condition.Maybe it is time that drug therapy should be considered.