<pre><FONT COLOR = BLACK>
<B>Modeling Alzheimer disease progression using composite cognitive test score
based on the data of NACC cohort </B>
   Lead Investigator:    Kirill Peskov
   Institution      :    Novartis Pharma
   E-Mail           :    kirill.peskov@novartis.com
   Proposal ID      :    430
</FONT>
<BR>
   Proposal Description:
   <textarea name="description" rows="12" cols="85" wrap="virtual" readonly>Recently, an empirically derived composite cognitive test score (CCTS) was developed for the investigation of cognitive decline in preclinical AD. This score showed improved power to detect early cognitive changes in the presenilin1 E280A mutation carriers [Ayutyanont N et al, (2014)] and cognitively normal elderly people [Langbaum JB et al., 2014]. Nevertheless, for the further application of CCTS for the clinical study design, e.g. as endpoint for the investigation of disease-modifying therapies, it is very important to understand how CCTS can be used for the characterization of continuous cognitive decline and search of the prognostic factors, which can differentiate individuals with higher progression rate from the others. This question can be addressed using application of disease progression modeling approach. Such studies have been done recently for the ADAS-cog clinical scores [Ito K et al., 2011 Samtani M et. al., 2012, 2013] and CDRSOB [Delor I et. al., 2013]. Thus, these modeling approaches are based on a concept that study entry doesn???t correspond to the start of the disease, and allow  reconstructing disease trajectory for a longer period than typically available in the study.  on the other side, development of the disease progression model can be used as a tool for  testing  different prognostic factors, e.g. Delor et al. found a set of biomarkers which can differentiate group of fast progressing mild cognitive impairment (MCI) patients with high risk of rapid development AD  symptoms. Accuracy to correctly predict MCI to AD transition according to this model was relatively high (more than 80) [Delor I et. al., 2013]. This fact evidences the positive value of these modeling methods for the improvement of the clinical diagnosis and prognosis of AD. Finally, by capturing the between  subject variability of the CCTS together with the typical disease progression and the covariates influencing it, these models allow to optimize trial designs of therapeuti </textarea>