Individual brain metabolic connectome indicator based on Kullback-Leibler Divergence Similarity Estimation predicts progression from mild cognitive impairment to Alzheimer\u2019s dementia

Decision letter: Stage-dependent differential influence of metabolic and structural networks on memory across Alzheimer’s disease continuum

Background: Changes in the metabolic and structural brain networks in mild cognitive impairment (MCI) have been widely researched. However, few studies have compared the differences in the topological properties of the metabolic and structural brain networks in patients with MCI.Methods: We analyzedmagnetic resonance imaging (MRI) and fluoro-deoxyglucose positron emission tomography (FDG-PET) data of 137 patients with MCI and 80 healthy controls (HCs). The HC group data comes from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database. The permutation test was used to compare the network parameters (characteristic path length, clustering coefficient, local efficiency, and global efficiency) between the two groups. Partial Pearson’s correlation analysis was used to calculate the correlations of the changes in gray matter volume and glucose intake in the key brain regions in MCI with the Alzheimer’s Disease Assessment Scale-Cognitive (ADAS-cog) sub-item scores.Results: Significant changes in the brain network parameters (longer characteristic path length, larger clustering coefficient, and lower local efficiency and global efficiency) were greater in the structural network than in the metabolic network (longer characteristic path length) in MCI patients than in HCs. We obtained the key brain regions (left globus pallidus, right calcarine fissure and its surrounding cortex, left lingual gyrus) by scanning the hubs. The volume of gray matter atrophy in the left globus pallidus was significantly positively correlated with comprehension of spoken language (p = 0.024) and word-finding difficulty in spontaneous speech item scores (p = 0.007) in the ADAS-cog. Glucose intake in the three key brain regions was significantly negatively correlated with remembering test instructions items in ADAS-cog (p = 0.020, p = 0.014, and p = 0.008, respectively).Conclusion: Structural brain networks showed more changes than metabolic brain networks in patients with MCI. Some brain regions with significant changes in betweenness centrality in both structural and metabolic networks were associated with MCI.

Background: Changes in the metabolic and structural brain networks in mild cognitive impairment (MCI) have been widely researched. However, few studies have compared the differences in the topological properties of the two brain networks assessed using magnetic resonance imaging (MRI) and fluoro-deoxyglucose positron emission tomography (FDG-PET) in patients with MCI. Methods: This study included 137 patients with MCI and 80 healthy controls (HCs). Sequential interictal scans were performed using FDG-PET and MRI. The MCI metabolic and structural brain networks were constructed according to the standardized uptake value ratio (SUVR) obtained using FDG-PET and gray matter volume obtained using MRI. The permutation test was used to compare the network parameters (characteristic path length, clustering coefficient, local efficiency, and global efficiency) between the two groups. Partial Pearson’s correlation analysis was used to calculate the correlations of the changes in gray matter volume and glucose intake in the key brain regions in MCI with the Alzheimer's Disease Assessment Scale-Cognitive (ADAS-cog) sub-item scores. Results: Significant changes in the brain network parameters (longer characteristic path length, larger clustering coefficient, and lower local efficiency and global efficiency) were greater in the structural network than in the metabolic network (longer characteristic path length) in MCI patients than in HCs. We obtained the key brain regions by scanning the hubs and found that the betweenness centrality of the right calcarine fissure and its surrounding cortex (CAL.R), left lingual gyrus (LING.L), and left globus pallidus (PAL.L) differed significantly between HCs and patients with MCI in both structural and metabolic networks (all p<0.05). The volume of gray matter atrophy in the PAL.L was significantly positively correlated with comprehension of spoken language (p=0.024) and word-finding difficulty in spontaneous speech item scores (p=0.007) in the ADAS-cog. Glucose intake in the three key brain regions (CAL.R, LING.L, and PAL.L) was significantly negatively correlated with remembering test instructions items in ADAS-cog (p=0.020, p=0.014, and p=0.008, respectively). Conclusion: MRI brain networks showed more changes than FDG-PET brain networks in patients with MCI. Some brain regions with significant changes in betweenness centrality in both structural and metabolic networks were associated with MCI.

249 Introduction: Cerebral β-amyloid load (Aβ) on one hand, and regional hypometabolism on the other hand were proposed as predictors of conversion from mild cognitive impairment (MCI) to Alzheimer’s dementia (AD). Given the limited availability of comparative studies, the present study examines the predictive values of F-18-Florbetapir (AV45), F-18-FDG (FDG) PET and clinical variables, separately and in combination in a large population. Methods: 319 MCI patients from the ADNI database (median follow-up: 47 [95% CI: 35-54] months) were studied. For FDG PET, we assessed the pattern expression score (PES) of a recently validated AD conversion-related pattern (ADCRP) [1], using voxel-based principal component analysis (PCA) [2]. For assessment of Aβ load with AV45 PET, we calculated the standardized uptake value ratio (SUVR) in AD-typical regions using the cerebellum as reference. In a training dataset (n=159), Cox proportional hazards regressions were applied to estimate the prognostic value of candidate predictors (adjusted for age and sex): i) clinical variables (APOE; functional activities questionnaire (FAQ)), ii) clinical variables combined with PES of ADCRP (FDG), and iii) clinical variables combined with Aβ load. For model validation, the results of each Cox regressions were applied to a test dataset (n=160) by calculating the prognostic index (PI) and stratifying each subject according to the predicted conversion risk (i.e., equally-sized groups of low-, medium- and high-risk). Results: PES (HR=2.38 per two standard deviations increase), FAQ (HR=2.12) and Aβ load (HR=2.09) were found to be significant independent predictors (all p≤0.001). In the training dataset, combining clinical variables with PES yielded a significantly better (p<0.001) model fit than combining clinical variables with Aβ or clinical variables alone (AIC=318, 327 and 339, respectively); best prediction accuracy was reached combining PES, Aβ and clinical variables into a combined model (AIC=300). 5-year conversion-free survival rates for the low, medium and high risk groups were 96%, 77% and 19% for PES, 97%, 64% and 44% for Aβ and 100%, 70% and 28% for PES and Aβ (clinical variables always included). Conclusions: Hypometabolism, Aβ and clinical variables represent complementary predictors of conversion from MCI to AD. The present study also supports the proposed NIA-AA research framework towards a biological definition of Alzheimer’s disease.

Synaptic dysfunction is linked to cognitive symptoms in Alzheimer's disease. Thus, measurement of synapse proteins in cerebrospinal fluid may be useful biomarkers to monitor synaptic degeneration. Cerebrospinal fluid levels of the postsynaptic protein neurogranin are increased in Alzheimer's disease, including in the predementia stage of the disease. Here, we tested the performance of cerebrospinal fluid neurogranin to predict cognitive decline and brain injury in the Alzheimer's Disease Neuroimaging Initiative study. An in-house immunoassay was used to analyse neurogranin in cerebrospinal fluid samples from a cohort of patients who at recruitment were diagnosed as having Alzheimer's disease with dementia (n = 95) or mild cognitive impairment (n = 173), as well as in cognitively normal subjects (n = 110). Patients with mild cognitive impairment were grouped into those that remained cognitively stable for at least 2 years (stable mild cognitive impairment) and those who progressed to Alzheimer's disease dementia during follow-up (progressive mild cognitive impairment). Correlations were tested between baseline cerebrospinal fluid neurogranin levels and baseline and longitudinal cognitive impairment, brain atrophy and glucose metabolism within each diagnostic group. Cerebrospinal fluid neurogranin was increased in patients with Alzheimer's disease dementia (P < 0.001), progressive mild cognitive impairment (P < 0.001) and stable mild cognitive impairment (P < 0.05) compared with controls, and in Alzheimer's disease dementia (P < 0.01) and progressive mild cognitive impairment (P < 0.05) compared with stable mild cognitive impairment. In the mild cognitive impairment group, high baseline cerebrospinal fluid neurogranin levels predicted cognitive decline as reflected by decreased Mini-Mental State Examination (P < 0.001) and increased Alzheimer's Disease Assessment Scale-cognitive subscale (P < 0.001) scores at clinical follow-up. In addition, high baseline cerebrospinal fluid neurogranin levels in the mild cognitive impairment group correlated with longitudinal reductions in cortical glucose metabolism (P < 0.001) and hippocampal volume (P < 0.001) at clinical follow-up. Furthermore, within the progressive mild cognitive impairment group, elevated cerebrospinal fluid neurogranin levels were associated with accelerated deterioration in Alzheimer's Disease Assessment Scale-cognitive subscale (β = 0.0017, P = 0.01). These data demonstrate that cerebrospinal fluid neurogranin is increased already at the early clinical stage of Alzheimer's disease and predicts cognitive deterioration and disease-associated changes in metabolic and structural biomarkers over time.

In population studies, most individuals with mild cognitive impairment (MCI) do not progress to dementia in the near term, but rather remain stable MCI or revert to normal cognition. Here, we characterized MCI subgroups with different outcomes over 5 years. A population-based cohort (N=1603). Clinical Dementia Rating (CDR); self-reported medical conditions, subjective cognitive concerns, self-rated health, depressive symptoms, blood pressure, medications, blood pressure, APOE genotype, cognitive domain composite scores. We compared 3 MCI subgroups who progressed to dementia (n=86), stabilized at MCI (n=384), or reverted to normal (n=252), to those who remained consistently normal (n=881), defining MCI as CDR = 0.5 and dementia as CDR≥1. Using multinomial logistic regression models adjusted for demographics, we examined the associations of each group with selected baseline characteristics. With the normal group for reference, worse subjective cognitive concerns, functional impairments, self-rated health, and depressive symptoms were associated with being in any MCI group. Taking more prescription medications was associated with being in the stable MCI and reverter groups; diabetes and low diastolic blood pressure were associated with stable MCI. The APOE4 genotype was associated with stable and progressive MCI; stroke was associated with progressive MCI. All MCI subgroups were likely to have lower mean composite scores in all cognitive domains and more operationally defined impairments in attention, language, and executive function; reverters were more likely to lack memory and visuospatial impairments. MCI subgroups with different 5-year outcomes had some distinct characteristics suggesting different underlying causes. The progressors, unlike the reverters, had a profile broadly typical of Alzheimer's disease; the stable MCIs had other, including vascular, morbidity. These data shed light on the heterogeneity of MCI in the population. J Am Geriatr Soc 67:232-238, 2019.

BackgroundAlzheimer’s disease‐related pattern (ADRP) is a metabolic brain biomarker of Alzheimer’s disease (AD), which was previously identified and validated. Its utility for differential diagnosis is known. Here, we aimed to explore its predictive value for conversion from cognitively normal (CN) to mild cognitive impairment (MCI), from MCI to dementia and for the rate of cognitive decline.MethodsWe analysed 609 participants followed for up to 15 years and their FDG PET scans from ADNI database (CN = 133, MCI = 352, dementia due to AD = 124) and 56 participants from University Medical centre Ljubljana (MCI = 22, AD = 34). ADRP expression was calculated from pre‐processed FDG PET scans. CN and MCI participants were stratified according to their conversion status and cerebrospinal fluid biomarkers. Amyloid positivity (A+) was defined as Aβ42 &lt; 880 pg/mL and tau positivity (T+) as pTau &gt; 21.8 pg/mL. AD patients were stratified into slow (&lt; 2 points/per year decline on mini mental state examination (MMSE)) and fast (&gt; 2 points/per year decline on MMSE) progressors.ResultsA+T+ cCN had higher baseline ADRP expression than A–T– (p&lt;0.001), A+T– (p = 0.003) and A+T+ (p = 0.01) stable CN corrected for age, while the groups did not differ in gender, education or baseline MMSE scores. A+T+ cMCI had higher baseline ADRP expression than A–T– (p&lt;0.001), A+T– (p = 0.001) and A+T+ (p&lt;0.001) stable MCI, corrected for baseline age and MMSE. A+T– stable MCI had higher baseline ADRP expression than A–T– stable MCI (p = 0.02). ADRP was a predictor of conversion to dementia in survival analysis (HR = 4.5[2.4‐8.7]). Similarly, in the UMCL dataset A+T+ cMCI had higher baseline ADRP expression than A–T– stable MCI (p = 0.03). Fast progressing AD patients had higher baseline ADRP expression than slowly progressing AD patients (p&lt;0.001). Similarly, in the UMCL dataset fast progressing AD patients had higher baseline ADRP expression than slowly progressing AD patients (p = 0.02). In neither dataset, AD groups differed in age, sex or baseline MMSE scores.ConclusionADRP expression levels can be used to predict conversion from CN to MCI and from MCI to dementia in A+T+ individuals. AD patients with high ADRP expression are at an elevated risk for faster progression of cognitive decline.

BackgroundAlzheimer’s disease‐related pattern (ADRP) is a metabolic brain biomarker of Alzheimer’s disease (AD), which was previously identified and validated. Its utility for differential diagnosis is known. Here, we aimed to explore its predictive value for conversion from cognitively normal (CN) to mild cognitive impairment (MCI), from MCI to dementia and for the rate of cognitive decline.MethodWe analysed 609 participants followed for up to 15 years and their FDG PET scans from ADNI database (CN = 133, MCI = 352, dementia due to AD = 124) and 56 participants from University Medical centre Ljubljana (MCI = 22, AD = 34). ADRP expression was calculated from pre‐processed FDG PET scans. CN and MCI participants were stratified according to their conversion status and cerebrospinal fluid biomarkers. Amyloid positivity (A+) was defined as Aß42 &lt; 880 pg/mL and tau positivity (T+) as pTau &gt; 21.8 pg/mL. AD patients were stratified into slow (&lt; 2 points/per year decline on mini mental state examination (MMSE)) and fast (&gt; 2 points/per year decline on MMSE) progressors.ResultA+T+ cCN had higher baseline ADRP expression than A–T– (p&lt;0.001), A+T– (p = 0.003) and A+T+ (p = 0.01) stable CN corrected for age, while the groups did not differ in gender, education or baseline MMSE scores. A+T+ cMCI had higher baseline ADRP expression than A–T– (p&lt;0.001), A+T– (p = 0.001) and A+T+ (p&lt;0.001) stable MCI, corrected for baseline age and MMSE. A+T– stable MCI had higher baseline ADRP expression than A–T– stable MCI (p = 0.02). ADRP was a predictor of conversion to dementia in survival analysis (HR = 4.5[2.4‐8.7]). Similarly, in the UMCL dataset A+T+ cMCI had higher baseline ADRP expression than A–T– stable MCI (p = 0.03). Fast progressing AD patients had higher baseline ADRP expression than slowly progressing AD patients (p&lt;0.001). Similarly, in the UMCL dataset fast progressing AD patients had higher baseline ADRP expression than slowly progressing AD patients (p = 0.02). In neither dataset, AD groups differed in age, sex or baseline MMSE scores.ConclusionADRP expression levels can be used to predict conversion from CN to MCI and from MCI to dementia in A+T+ individuals. AD patients with high ADRP expression are at an elevated risk for faster progression of cognitive decline.

Most current research on Alzheimer’s disease (AD) is based on transverse measurements. Given the nature of neurodegeneration in AD progression, observing longitudinal changes in the structural features of brain networks over time may improve the accuracy of the predicted transformation and provide a good measure of the progression of AD. Currently, there is no cure for patients with existing AD dementia, but patients with mild cognitive impairment (MCI) in the prodromal stage of AD dementia may be diagnosed. The study of the early diagnosis of MCI and the prediction of MCI to AD transformation is of great significance for the monitoring of the MCI to AD transformation process. Despite the high rate of MCI conversion to AD, the neuropathological cause of MCI is heterogeneous. However, many people with MCI remain stable. Treatment options are different for patients with stable MCI and those with underlying dementia. Therefore, it is of great significance for clinical practice to predict whether patients with MCI will develop AD dementia. This paper proposes an improved algorithm that is based on a convolution neural network (CNN) with residuals combined with multi-layer long short-term memory (LSTM) to diagnose AD and predict MCI. Firstly, multi-time resting-state fMRI images were obtained from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database for preprocessing, and then an AAL brain partition template was used to construct a 90 × 90 functional connectivity (FC) network matrix of a whole-brain region of interest (ROI). Secondly, the diversity of training samples was increased by generating an adversarial network (GAN). Finally, a CNN with residuals and a multi-layer LSTM model were constructed to automatically classify and predict the functional adjacency matrix. This method can not only distinguish Alzheimer’s disease from normal health conditions at multiple time points, but can also predict progressive MCI (pMCI) and stable MCI (sMCI) at multiple time points. The classification accuracies in AD vs. NC and sMCI vs.pMCI reached 93.5% and 75.5%, respectively.

ID 114 – Progressive MCI proved with vMRI shows lower delta ERO than stable MCI at baseline: A longitudinal study

BackgroundControversy existed in the prognosis of reversion from mild cognitive impairment (MCI) to normal cognition (NC). We aim to depict the prognostic characteristics of cognition, neuroimaging, and pathology biomarkers, as well as the risk of Alzheimer’s disease (AD) dementia for MCI reverters.MethodsA total of 796 non-demented participants (mean age = 73.3 years, female = 54.4%, MCI = 55.7%), who were divided into MCI reverters (n = 109), stable MCI (n = 334), and stable NC (n = 353) based on 2-year diagnosis changes, were subsequently followed up for 6 years. Cox proportional hazard regression models were applied to assess the AD dementia hazard. Linear mixed-effect models were used to evaluate the differences in changing rates of cognitive scores, brain volumes, brain metabolism, and AD biomarkers among three groups.ResultsThe 2-year MCI reversion rate was 18.17%. MCI reversion was associated with an 89.6% lower risk of AD dementia (HR = 0.104, 95% confidence interval = [0.033, 0.335], p < 0.001) than stable MCI. No significant difference in incident AD risk was found between MCI reverters and stable NC (p = 0.533). Compared to stable MCI, reverters exhibited slower decreases in cognitive performance (false discovery rate corrected p value [FDR-Q] < 0.050), brain volumes (FDR-Q < 0.050), brain metabolism (FDR-Q < 0.001), and levels of cerebrospinal fluid β-amyloid1–42 (FDR-Q = 0.008). The above-mentioned differences were not found between MCI reverters and stable NC (FDR-Q > 0.050).ConclusionsReversion from MCI to NC predicts a favorable prognosis of pathological, neurodegenerative, and cognitive trajectory.

Prediction of Alzheimer's disease dementia with MRI beyond the short-term: Implications for the design of predictive models.

Morphological brain network plays a key role in investigating abnormalities in neurological diseases such as mild cognitive impairment (MCI) and Alzheimer's disease (AD). However, most of the morphological brain network construction methods only considered a single morphological feature. Each type of morphological feature has specific neurological and genetic underpinnings. A combination of morphological features has been proven to have better diagnostic performance compared with a single feature, which suggests that an individual morphological brain network based on multiple morphological features would be beneficial in disease diagnosis. Here, we proposed a novel method to construct individual morphological brain networks for two datasets by calculating the exponential function of multivariate Euclidean distance as the evaluation of similarity between two regions. The first dataset included 24 healthy subjects who were scanned twice within a 3-month period. The topological properties of these brain networks were analyzed and compared with previous studies that used different methods and modalities. Small world property was observed in all of the subjects, and the high reproducibility indicated the robustness of our method. The second dataset included 170 patients with MCI (86 stable MCI and 84 progressive MCI cases) and 169 normal controls (NC). The edge features extracted from the individual morphological brain networks were used to distinguish MCI from NC and separate MCI subgroups (progressive vs. stable) through the support vector machine in order to validate our method. The results showed that our method achieved an accuracy of 79.65% (MCI vs. NC) and 70.59% (stable MCI vs. progressive MCI) in a one-dimension situation. In a multiple-dimension situation, our method improved the classification performance with an accuracy of 80.53% (MCI vs. NC) and 77.06% (stable MCI vs. progressive MCI) compared with the method using a single feature. The results indicated that our method could effectively construct an individual morphological brain network based on multiple morphological features and could accurately discriminate MCI from NC and stable MCI from progressive MCI, and may provide a valuable tool for the investigation of individual morphological brain networks.

A variety of imaging, neuropsychological, and genetic biomarkers have been suggested as potential biomarkers for the identification of mild cognitive impairment (MCI) in patients who later develop Alzheimer's disease (AD). Here, we systematically evaluated the most promising combinations of these biomarkers regarding discrimination between stable and converter MCI and reflection of disease staging. Alzheimer's Disease Neuroimaging Initiative data of AD (n = 144), controls (n = 112), stable (n = 265) and converter (n = 177) MCI, for which apolipoprotein E status, neuropsychological evaluation, and structural, glucose, and amyloid imaging were available, were included in this study. Naïve Bayes classifiers were built on AD and controls data for all possible combinations of these biomarkers, with and without stratification by amyloid status. All classifiers were then applied to the MCI cohorts. We obtained an accuracy of 76% for discrimination between converter and stable MCI with glucose positron emission tomography as a single biomarker. This accuracy increased to about 87% when including further imaging modalities and genetic information. We also identified several biomarker combinations as strong predictors of time to conversion. Use of amyloid validated training data resulted in increased sensitivities and decreased specificities for differentiation between stable and converter MCI when amyloid was included as a biomarker but not for other classifier combinations. Our results indicate that fully independent classifiers built only on AD and controls data and combining imaging, genetic, and/or neuropsychological biomarkers can more reliably discriminate between stable and converter MCI than single modality classifiers. Several biomarker combinations are identified as strongly predictive for the time to conversion to AD.

Alzheimer's disease (AD) is characterised by a dynamic process of neurocognitive changes from normal cognition to mild cognitive impairment (MCI) and progression to dementia. However, not all individuals with MCI develop dementia. Predicting whether individuals with MCI will decline (i.e. progressive MCI) or remain stable (i.e. stable MCI) is impeded by patient heterogeneity due to comorbidities that may lead to MCI diagnosis without progression to AD. Despite the importance of early diagnosis of AD for prognosis and personalised interventions, we still lack robust tools for predicting individual progression to dementia. Here, we propose a novel trajectory modelling approach based on metric learning (Generalised Metric Learning Vector Quantization) that mines multimodal data from MCI patients in the Alzheimer's disease Neuroimaging Initiative (ADNI) cohort to derive individualised prognostic scores of cognitive decline due to AD. We develop an integrated biomarker generation- using partial least squares regression- and classification methodology that extends beyond binary patient classification into discrete subgroups (i.e. stable vs. progressive MCI), determines individual profiles from baseline (i.e. cognitive or biological) data and predicts individual cognitive trajectories (i.e. change in memory scores from baseline). We demonstrate that a metric learning model trained on baseline cognitive data (memory, executive function, affective measurements) discriminates stable vs. progressive MCI individuals with high accuracy (81.4%), revealing an interaction between cognitive (memory, executive functions) and affective scores that may relate to MCI comorbidity (e.g. affective disturbance). Training the model to perform the same binary classification on biological data (mean cortical β-amyloid burden, grey matter density, APOE 4) results in similar prediction accuracy (81.9%). Extending beyond binary classifications, we develop and implement a trajectory modelling approach that shows significantly better performance in predicting individualised rate of future cognitive decline (i.e. change in memory scores from baseline), when the metric learning model is trained with biological (r=-0.68) compared to cognitive (r=-0.4) data. Our trajectory modelling approach reveals interpretable and interoperable markers of progression to AD and has strong potential to guide effective stratification of individuals based on prognostic disease trajectories, reducing MCI patient misclassification, that is critical for clinical practice and discovery of personalised interventions.