Accuracy of biomarkers in predicting conversion to dementia in MCI subjects: The statistical analysis plan of the interceptor study

Background and aims: Many clinical trials on potentially disease-modifying drugs are focused to mild cognitive impairment (MCI) prodromal-to-Alzheimer's disease. The MCI population actually includes patients with higher risk of progression to AD. Since the new drugs, if effective, will carry both elevated unit costs and not marginal side effects, they should be administered selectively to high-risk subjects. The Italian INTERCEPTOR project, promoted by the Italian Medicine Agency and the Italian Ministry of Health, is a multicenter, interventional, nontherapeutic cohort study in subjects with MCI, with the primary aim of identifying biomarkers that better predict the conversion to AD after 3 years of follow-up. Methods: A sample of 500 subjects with MCI was planned to be enrolled. The biomarkers under investigation obtained by the analysis of MMSE, CSF, FDG-PET, FCRST, APOE4, EEG, Volumetric MRI. Multivariate prediction model will provide the predictive performance of each biomarker and combinations. In applying biomarkers, three scenarios are considered: 1) use of cut-offs indicated by experts 2) defining new optimal cut-offs for the specific population 3) use of values in a continuous form in order to evaluate all predictive information. Finally, nomograms will be defined for use in clinical practice. Results: Due to COVID-19, the target sample size was not reached. The enrolment was closed on October 31st 2020. Overall 498 patients were screened, 362 recruited into the study and 17 dropped out. Conclusions: The findings will support the diagnostic path redefinition to identify those patients in the early stage eligible to prescription for disease-modifying medications.