John Tuje Ikhena MPH, Presenter: Nothing to Disclose

Bharath Gopal Rathakrishnan BS, Abstract Co-Author: Nothing to Disclose

P. Murali Doraiswamy MD, Abstract Co-Author: Research Consultant, Bristol-Myers Squibb Company Research Consultant, Eli Lilly and Company Research Consultant, Neuronetrix, Inc Research Consultant, Medivation, Inc Research Grant, Bristol-Myers Squibb Company Research Grant, Eli Lilly and Company Research Grant, Neuronetrix, Inc Research Grant, Medivation, Inc Stockholder, Sonexa Therapeutics, Inc Stockholder, Clarimedix, Inc Speaker, Forest Medical, LLC

Jeffrey Robert Petrella MD, Abstract Co-Author: Advisory Board, Johnson & Johnson Speakers Bureau, Quintiles Inc Advisory Board, Piramal Enterprises Limited

To retrospectively use NeuroQuant to assess how well age-adjusted volumetric measures perform in predicting conversion to Alzheimer’s disease in patients with mild cognitive impairment (MCI).

We selected data for subjects from the Alzheimer’s disease Neuroimaging Initiative (ADNI) database. Participating study sites were approved by their respective institutional review boards and all subjects provided full written informed consents prior to data collection. All protected health information from the patient data was de-identified. We analyzed data from 281 subjects with mild cognitive impairment (260 with late MCI; 34% female) in the ADNI database and assessed the predictive values of MMSE, hippocampal volumes and lateral ventricles volume in converting to Alzheimer’s disease over a 3-year follow-up period. MRI volumetrics were derived from T1-weighted magnetic resonance (MR) imaging data from NeuroQuant. Data obtained were analyzed using Chi-Square test, Receiver operating characteristic (ROC) analysis and regression models.

46% of patients with late MCI converted to Alzheimer’s (110 subjects total) at 3 years follow up. We found that hippocampal volume has a 69.5% likelihood of predicting conversion to Alzheimer’s (AUC – 0.695, p<0.05). MMSE and HV showed a significant effect in predicting conversion to Alzheimer’s (AUC – 0.74, p<0.05). Lateral ventricle volumes of subjects were not significantly associated predicting conversion to Alzheimer’s disease (AUC – 0.518, p = 0.82).

Among the various age-adjusted NeuroQuant measures we analyzed, hippocampal volume was found to be the most sensitive in predicting conversion to Alzheimer’s in MCI subjects. Sensitivity increased when MMSE was added to these estimates. Therefore we conclude that in developing a predictive model, it would be vital to include MMSE and hippocampal volume of subjects.

Our study quantitatively examines the utility of a currently available clinical implementation of automated volumetric assessment software (NeuroQuant) in the evaluation of MRI scans for predicting the conversion of MCI patients to Alzheimer’s disease. We will use this information to create an individualized risk-of-conversion profile for individual MCI patient’s based on volumetrics and other readily available clinical formation.

Ikhena, J, Rathakrishnan, B, Doraiswamy, P, Petrella, J, Utility of Automated MRI Brain Volumetrics in Predicting Conversion of Mild Cognitive Impairment to Alzheimer’s Disease: A Retrospective Study in the De-identified National ADNI Database.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14011085.html