Combining Polygenic Hazard Score With Volumetric MRI and Cognitive Measures Improves Prediction of Progression From Mild Cognitive Impairment to Alzheimer's Disease.

Karolina Kauppi,Rahul S Desikan,Dominic Holland,Anders M Dale,Linda K Mcevoy,Chin Hong Tan,Ole A Andreassen,Chun Chieh Fan,Chi-Hua Chen

doi:10.3389/fnins.2018.00260

Abstract

Improved prediction of progression to Alzheimer's Disease (AD) among older individuals with mild cognitive impairment (MCI) is of high clinical and societal importance. We recently developed a polygenic hazard score (PHS) that predicted age of AD onset above and beyond APOE. Here, we used data from the Alzheimer's Disease Neuroimaging Initiative (ADNI) to further explore the potential clinical utility of PHS for predicting AD development in older adults with MCI. We examined the predictive value of PHS alone and in combination with baseline structural magnetic resonance imaging (MRI) data on performance on the Mini-Mental State Exam (MMSE). In survival analyses, PHS significantly predicted time to progression from MCI to AD over 120 months (p = 1.07e-5), and PHS was significantly more predictive than APOE alone (p = 0.015). Combining PHS with baseline brain atrophy score and/or MMSE score significantly improved prediction compared to models without PHS (three-factor model p = 4.28e-17). Prediction model accuracies, sensitivities and area under the curve were also improved by including PHS in the model, compared to only using atrophy score and MMSE. Further, using linear mixed-effect modeling, PHS improved the prediction of change in the Clinical Dementia Rating—Sum of Boxes (CDR-SB) score and MMSE over 36 months in patients with MCI at baseline, beyond both APOE and baseline levels of brain atrophy. These results illustrate the potential clinical utility of PHS for assessment of risk for AD progression among individuals with MCI both alone, or in conjunction with clinical measures of prodromal disease including measures of cognitive function and regional brain atrophy.

Highlights

Late onset Alzheimer’s disease (AD) is the most common form of dementia, affecting 24–35 million people world-wide (Querfurth and LaFerla, 2010; Alzheimer’s Association, 2015)
While most studies examining genetic risk for sporadic AD focus on Apolipoprotein E (APOE) genotype, some have assessed polygenic risk scores beyond APOE based on case-control data from large-scale genome-wide association studies (GWAS) (Lambert et al, 2013), and found that genetic variants associated with elevated AD risk influence brain structure (Sabuncu et al, 2012) and cognitive function (Marioni et al, 2017)
We investigated the clinical utility of polygenic hazard score (PHS) for individual assessment of risk for clinical progression to AD over time among older individuals with mild cognitive impairment (MCI), a critical question for most patients admitted to memory clinics

Summary

Introduction

Late onset Alzheimer’s disease (AD) is the most common form of dementia, affecting 24–35 million people world-wide (Querfurth and LaFerla, 2010; Alzheimer’s Association, 2015). Inheritance of the ε4 allele of the Apolipoprotein E (APOE) gene is the strongest individual risk factor for late onset AD (Yu et al, 2014). Given that the incidence of AD increases sharply with age, we recently develop a polygenic hazard score (PHS) for prediction of age-specific AD risk, based on 31 AD-susceptibility variants, including APOE (Desikan et al, 2017). The PHS showed substantial improvement over APOE in predicting age of AD onset and was associated with biomarkers of AD, including MRIbased hippocampal volume loss (Desikan et al, 2017), amyloid, and tau deposition (Tan et al, 2018)

Methods

Results

Conclusion