Generation and characterization of an iPS cell line (PUMCi006-A) from skin fibroblasts of a patient with an M239T mutation in PSEN2 gene.

Presenilin-2 (PSEN2) mutation is one of the pathogenic factors of autosomal dominant early-onset Alzheimer's disease (EOAD). We generated a human induced pluripotent stem cell (iPSC) line from fibroblasts of an EOAD patient carrying PSEN2 mutation (c.716 T > C) utilizing Sendai reprogramming kit. The resulting iPSC line carried patient-specific point mutation, exhibited typical iPSC morphology, retained a normal karyotype, expressed pluripotency markers, and could form embryoid bodies. Established iPSC line serve as valuable resource for EOAD disease pathogenesis modelling and drug screening.

GRN mutation spectrum and genotype-phenotype correlation in Chinese dementia patients: data from PUMCH dementia cohort.

BACKGROUND: METHODS: The GRN mutations, especially of the loss of function type, are causative of frontotemporal dementia (FTD). However, several GRN variants can be found in other neurodegenerative diseases, such as Alzheimer's disease (AD) and Parkinson's disease. So far, there have been over 300 GRN mutations reported globally. However, the genetic spectrum and phenotypic characteristics have not been fully elucidated in Chinese population.The participants were from the dementia cohort of Peking Union Medical College Hospital (n=1945). They received history inquiry, cognitive evaluation, brain imaging and exome sequencing. The dementia subjects carrying the rare variants of the GRN were included in this study. Those with the pathogenic or likely pathogenic variants of other dementia-related genes were excluded. RESULTS: 14 subjects carried the rare variants of GRN. They were clinically diagnosed with behavioural variant of FTD (n=2), non-fluent/agrammatic variant primary progressive aphasia (PPA, n=3), semantic variant PPA (n=1), AD (n=6) and mixed dementia (n=2). 13 rare variants of GRN were found, including 6 novel variants (W49X, S226G, M152I, A91E, G79E and A303S). The most prevalent symptom was amnesia (85.7%, 12/14), followed by psychiatric and behavioural disorder (78.6%, 11/14). In terms of lobar atrophy, temporal atrophy/hypometabolism was the most common (85.7%, 12/14), followed by parietal atrophy/hypometabolism (78.6%, 11/14). CONCLUSION: The novel GRN variants identified in this study contribute to enrich the GRN mutation repertoire. There is phenotypic similarity and diversity among Chinese patients with the GRN mutations.

Comorbidity of Dementia: A Cross-Sectional Study of PUMCH Dementia Cohort.

BACKGROUND: Comorbidities reduce quality of life for people with dementia and caregivers. Some comorbidities share a genetic basis with dementia. OBJECTIVE: The objective of this study is to assess comorbidity in patients with different dementia subtypes in order to better understand the pathogenesis of dementias. METHODS: A total of 298 patients with dementia were included. We collected some common comorbidities. We analyzed the differences in comorbidities among patients with dementia according to clinical diagnosis, age of onset (early-onset: < 65 and late-onset: ≥65 years old) and apolipoprotein (APOE) genotypes by using the univariate and multivariate approaches. RESULTS: Among 298 participants, there were 183 Alzheimer's disease (AD), 40 vascular dementia (VaD), 37 frontotemporal dementia (FTLD), 20 Lewy body dementia (LBD), and 18 other types of dementia. Based on age of onset, 156 cases had early-onset dementia and 142 cases had late-onset dementia. The most common comorbidities observed in all dementia patients were hyperlipidemia (68.1%), hypertension (39.9%), insomnia (21.1%), diabetes mellitus (19.5%), and hearing impairment (18.1%). The prevalence of hypertension and cerebrovascular disease was found to be higher in patients with VaD compared to those with AD (p = 0.002, p < 0.001, respectively) and FTLD (p = 0.028, p = 0.004, respectively). Additionally, patients with late-onset dementia had a higher burden of comorbidities compared to those with early-onset dementia. It was observed that APOE ɛ4/ɛ4 carriers were less likely to have insomnia (p = 0.031). CONCLUSIONS: Comorbidities are prevalent in patients with dementia, with hyperlipidemia, hypertension, insomnia, diabetes, and hearing impairment being the most commonly observed. Comorbidity differences existed among different dementia subtypes.

Assessment of cheese sign and its association with vascular risk factors: Data from PUMCH dementia cohort.

BACKGROUND: In the clinic, practitioners encounter many patients with an abnormal pattern of dense punctate magnetic resonance imaging (MRI) signal in the basal ganglia, a phenomenon known as "cheese sign". This sign is reported as common in cerebrovascular diseases, dementia, and old age. Recently, cheese sign has been speculated to consist of dense perivascular space (PVS). This study aimed to assess the lesion types of cheese sign and analyze the correlation between this sign and vascular disease risk factors. METHODS: A total of 812 patients from Peking Union Medical College Hospital (PUMCH) dementia cohort were enrolled. We analyzed the relationship between cheese sign and vascular risk. For assessing cheese sign and defining its degree, the abnormal punctate signals were classified into basal ganglia hyperintensity (BGH), PVS, lacunae/infarctions and microbleeds, and counted separately. Each type of lesion was rated on a four-level scale, and then the sum was calculated; this total was defined as the cheese sign score. Fazekas and Age-Related White Matter Changes (ARWMC) scores were used to evaluate the paraventricular, deep, and subcortical gray/white matter hyperintensities. RESULTS: A total of 118 patients (14.5%) in this dementia cohort were found to have cheese sign. Age (odds ratio [OR]: 1.090, 95% confidence interval [CI]: 1.064-1.120, P <0.001), hypertension (OR: 1.828, 95% CI: 1.123-2.983, P = 0.014), and stroke (OR: 1.901, 95% CI: 1.092-3.259, P = 0.025) were risk factors for cheese sign. There was no significant relationship between diabetes, hyperlipidemia, and cheese sign. The main components of cheese sign were BGH, PVS, and lacunae/infarction. The proportion of PVS increased with cheese sign severity. CONCLUSIONS: The risk factors for cheese sign were hypertension, age, and stroke. Cheese sign consists of BGH, PVS, and lacunae/infarction.

Association Between Wnt Target Genes and Cortical Volumes in Alzheimer's Disease.

The disproportionate cortical atrophy is an established biomarker for the pathophysiological process of Alzheimer's disease (AD). However, the genetic basis underlying the cortical atrophy remains poorly defined. Herein, we aim to illustrate the effect of the Wnt target genes on the cortical volumes of AD patients. 82 sporadic AD patients were recruited. All the subjects had history survey, blood biochemical examination, cognitive assessment, MRI morphometry and whole exome sequencing. This report focused on 84 common variants (minor allele frequency > 0.01) of 32 Wnt target genes, including the APC, DAAM1, DACT1, DISC1, LATS2, TLR2, WDR61, and the AXIN, DVL, FZD, LRP, TCF/LEF, WNT family genes. The Wnt target genes showed asymmetric effects on the cortical volumes of AD patients. The right temporal/parietal/occipital cortices were more affected than left temporal/parietal/occipital cortices. Nevertheless, the reverse applied to the frontal cortex. The DACT1 affected the cortical thickness most, followed by the TCF3 and APC. The DACT1 rs698025-GG genotype displayed greater right temporal pole and left medial orbito-frontal gyrus than rs698025-GA genotype (2.4 ± 0.4 vs. 2.0 ± 0.6, P = 0.005; 5.2 ± 0.6 vs. 5.0 ± 0.6, P = 0.001). The brain region most influenced by the Wnt target genes was the right calcarine cortex. In conclusion, the common variants of the Wnt target genes exert asymmetric effects on the cortical volumes of AD patients. The Wnt signaling pathway may play a role in the cortical atrophy of AD patients.

Genotypic and phenotypic heterogeneity among Chinese pediatric genetic white matter disorders.

BACKGROUND: The pediatric genetic white matter disorders are characterized by a broad disease spectrum. Genetic testing is valuable in the diagnosis. However, there are few studies on the clinical and genetic spectrum of Chinese pediatric genetic white matter disorders. METHODS: The participants were enrolled from the cohort of Peking Union Medical College Hospital. They all received history collection, brain MRI and gene sequencing. Their neurologic complaints which were related to white matter disorders occurred before 18. Brain MRI indicated periventricular and/or deep white matter lesions, fazekas grade 2-3. RESULTS: Among the 13 subjects, there were 11 males and two females. The average age of onset was 10.0 ± 5.5 years old. The potential genetic variants were found in 84.6% (11/13) subjects. The ABCD1 showed the greatest mutation frequency (30.8%, 4/13). The EIF2B3 A151fs, EIF2B4 c.885 + 2T > G, EIF2B5 R129X and MPV17 Q142X were novel pathogenic/likely pathogenic variants. 100% (4/4) ABCD1 carriers were accompanied by visual impairment, whereas 100% (3/3) EIF2B carriers developed dysuria. 100% (4/4) ABCD1 carriers exhibited diffuse white matter hyperintensities mainly in the posterior cortical regions, while the EIF2B4 and EIF2B5 carriers were accompanied by cystic degeneration. CONCLUSION: There is genotypic and phenotypic heterogeneity among Chinese subjects with pediatric genetic white matter disorders. The knowledge of these clinical and genetic characteristics facilitates an accurate diagnosis of these diseases.

Toward a stable and low-resource PLM-based medical diagnostic system via prompt tuning and MoE structure.

Machine learning (ML) has been extensively involved in assistant disease diagnosis and prediction systems to emancipate the serious dependence on medical resources and improve healthcare quality. Moreover, with the booming of pre-training language models (PLMs), the application prospect and promotion potential of machine learning methods in the relevant field have been further inspired. PLMs have recently achieved tremendous success in diverse text processing tasks, whereas limited by the significant semantic gap between the pre-training corpus and the structured electronic health records (EHRs), PLMs cannot converge to anticipated disease diagnosis and prediction results. Unfortunately, establishing connections between PLMs and EHRs typically requires the extraction of curated predictor variables from structured EHR resources, which is tedious and labor-intensive, and even discards vast implicit information.In this work, we propose an Input Prompting and Discriminative language model with the Mixture-of-experts framework (IPDM) by promoting the model's capabilities to learn knowledge from heterogeneous information and facilitating the feature-aware ability of the model. Furthermore, leveraging the prompt-tuning mechanism, IPDM can inherit the impacts of the pre-training in downstream tasks exclusively through minor modifications. IPDM remarkably outperforms existing models, proved by experiments on one disease diagnosis task and two disease prediction tasks. Finally, experiments with few-feature and few-sample demonstrate that IPDM achieves significant stability and impressive performance in predicting chronic diseases with unclear early-onset characteristics or sudden diseases with insufficient data, which verifies the superiority of IPDM over existing mainstream methods, and reveals the IPDM can powerfully address the aforementioned challenges via establishing a stable and low-resource medical diagnostic system for various clinical scenarios.

Association of APOE ε4/ε4 with fluid biomarkers in patients from the PUMCH dementia cohort.

Background: Apolipoprotein-E (APOE) ε4 is a major genetic risk factor for Alzheimer's disease (AD). Current studies, which were mainly based on the clinical diagnosis rather than biomarkers, come to inconsistent conclusions regarding the associations of APOE ε4 homozygotes (APOE ε4/ε4) and cerebrospinal fluid (CSF) biomarkers of AD. In addition, few studies have explored the associations of APOE ε4/ε4 with plasma biomarkers. Therefore, we aimed to investigate the associations of APOE ε4/ε4 with fluid biomarkers in dementia and biomarker-diagnosed AD. Methods: A total of 297 patients were enrolled. They were classified into Alzheimer's continuum, AD, and non-AD, according to CSF biomarkers and/or ß amyloid PET results. AD was a subgroup of the AD continuum. Plasma Amyloid ß (Aß) 40, Aß42, glial fibrillary acidic protein (GFAP), neurofilament light chain (NFL), and phosphorylated tau (P-tau)181 were quantified in 144 of the total population using an ultra-sensitive Simoa technology. We analyzed the associations of APOE ε4/ε4 on CSF and plasma biomarkers in dementia and biomarker diagnosed AD. Results: Based on the biomarker diagnostic criteria, 169 participants were diagnosed with Alzheimer's continuum and 128 individuals with non-AD, and among the former, 120 patients with AD. The APOE ε4/ε4 frequencies were 11.8% (20/169), 14.2% (17/120), and 0.8% (1/128) in Alzheimer's continuum, AD and non-AD, respectively. Only CSF Aß42 was shown to be decreased in APOE ε4/ε4 carriers than in non-carriers for patients with AD (p = 0.024). Furthermore, we did not find any associations of APOE ε4 with plasma biomarkers of AD and non-AD. Interestingly, we found that in non-AD patients, APOE ε4 carriers had lower CSF Aß42 (p = 0.018) and higher T-tau/Aß42 ratios (p < 0.001) and P-tau181/Aß42 ratios (p = 0.002) than non-carriers. Conclusion: Our data confirmed that of the three groups (AD continuum, AD, and non-AD), those with AD had the highest frequency of APOE ɛ4/ɛ4 genotypes. The APOE ɛ4/ɛ4 was associated with CSF levels of Aß42 but not tau for AD and non-AD, suggesting that APOE ɛ4/ɛ4 affected the Aß metabolism of both. No associations between APOE ε4/ɛ4 and plasma biomarkers of AD and non-AD were found.

Distribution of Cortical Atrophy Associated with Cognitive Decline in Alzheimer's Disease: A Cross-Sectional Quantitative Structural MRI Study from PUMCH Dementia Cohort.

BACKGROUND: Quantitative measures of atrophy on structural MRI are sensitive to the neurodegeneration that occurs in AD, and the topographical pattern of atrophy could serve as a sensitive and specific biomarker. OBJECTIVE: We aimed to examine the distribution of cortical atrophy associated with cognitive decline and disease stage based on quantitative structural MRI analysis in a Chinese cohort to inform clinical diagnosis and follow-up of AD patients. METHODS: One hundred and eleven patients who were clinically diagnosed with probable AD were enrolled. All patients completed a systemic cognitive evaluation and domain-specific batteries. The severity of cognitive decline was defined by MMSE score: 1-10 severe, 11-20 moderate, and 21-30 mild. Cortical volume and thickness determined using 3D-T1 MRI data were analyzed using voxelbased morphometry and surface-based analysis supported by the DR. Brain Platform. RESULTS: The male:female ratio was 38:73. The average age was 70.8 ± 10.6 years. The mild: moderate: severe ratio was 48:38:25. Total grey matter volume was significantly related to cognition while the relationship between white matter volume and cognition did not reach statistical significance. The volume of the temporal-parietal-occipital cortex was most strongly associated with cognitive decline in group analysis, while the hippocampus and entorhinal area had a less significant association with cognitive decline. Volume of subcortical grey matter was also associated with cognition. Volume and thickness of temporoparietal cortexes were significantly correlated with the cognitive decline, with a left predominance observed. CONCLUSION: Cognitive deterioration was associated with cortical atrophy. Volume and thickness of the left temporal-parietal-occipital cortex were most important in early diagnosis and longitudinal evaluation of AD in clinical practice. Cognitively relevant cortices were left predominant.

Clinical features of NOTCH2NLC-related neuronal intranuclear inclusion disease.

BACKGROUND: Abnormal expanded GGC repeats within the NOTCH2HLC gene has been confirmed as the genetic mechanism for most Asian patients with neuronal intranuclear inclusion disease (NIID). This cross-sectional observational study aimed to characterise the clinical features of NOTCH2NLC-related NIID in China. METHODS: Patients with NOTCH2NLC-related NIID underwent an evaluation of clinical symptoms, a neuropsychological assessment, electrophysiological examination, MRI and skin biopsy. RESULTS: In the 247 patients with NOTCH2NLC-related NIID, 149 cases were sporadic, while 98 had a positive family history. The most common manifestations were paroxysmal symptoms (66.8%), autonomic dysfunction (64.0%), movement disorders (50.2%), cognitive impairment (49.4%) and muscle weakness (30.8%). Based on the initial presentation and main symptomology, NIID was divided into four subgroups: dementia dominant (n=94), movement disorder dominant (n=63), paroxysmal symptom dominant (n=61) and muscle weakness dominant (n=29). Clinical (42.7%) and subclinical (49.1%) peripheral neuropathies were common in all types. Typical diffusion-weighted imaging subcortical lace signs were more frequent in patients with dementia (93.9%) and paroxysmal symptoms types (94.9%) than in those with muscle weakness (50.0%) and movement disorders types (86.4%). GGC repeat sizes were negatively correlated with age of onset (r=-0.196, p<0.05), and in the muscle weakness-dominant type (median 155.00), the number of repeats was much higher than in the other three groups (p<0.05). In NIID pedigrees, significant genetic anticipation was observed (p<0.05) without repeat instability (p=0.454) during transmission. CONCLUSIONS: NIID is not rare; however, it is usually misdiagnosed as other diseases. Our results help to extend the known clinical spectrum of NOTCH2NLC-related NIID.

Learning Cognitive-Test-Based Interpretable Rules for Prediction and Early Diagnosis of Dementia Using Neural Networks.

BACKGROUND: Accurate, cheap, and easy to promote methods for dementia prediction and early diagnosis are urgently needed in low- and middle-income countries. Integrating various cognitive tests using machine learning provides promising solutions. However, most effective machine learning models are black-box models that are hard to understand for doctors and could hide potential biases and risks. OBJECTIVE: To apply cognitive-test-based machine learning models in practical dementia prediction and diagnosis by ensuring both interpretability and accuracy. METHODS: We design a framework adopting Rule-based Representation Learner (RRL) to build interpretable diagnostic rules based on the cognitive tests selected by doctors. According to the visualization and test results, doctors can easily select the final rules after analysis and trade-off. Our framework is verified on the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset (n = 606) and Peking Union Medical College Hospital (PUMCH) dataset (n = 375). RESULTS: The predictive or diagnostic rules learned by RRL offer a better trade-off between accuracy and model interpretability than other representative machine learning models. For mild cognitive impairment (MCI) conversion prediction, the cognitive-test-based rules achieve an average area under the curve (AUC) of 0.904 on ADNI. For dementia diagnosis on subjects with a normal Mini-Mental State Exam (MMSE) score, the learned rules achieve an AUC of 0.863 on PUMCH. The visualization analyses also verify the good interpretability of the learned rules. CONCLUSION: With the help of doctors and RRL, we can obtain predictive and diagnostic rules for dementia with high accuracy and good interpretability even if only cognitive tests are used.

Genetic Spectrum and Clinical Heterogeneity of Chinese Frontotemporal Dementia Patients: Data from PUMCH Dementia Cohort.

BACKGROUND: There are relatively few data on the genetic spectrum of Chinese frontotemporal dementia (FTD) population. OBJECTIVE: With the dementia cohort of Peking Union Medical College Hospital, we aim to illustrate the genetic spectrum of FTD patients, as well as the phenotypic heterogeneity of FTD-gene variant carriers. METHODS: 204 unrelated, clinically diagnosed FTD patients of Chinese ancestry were enrolled. All the participants received demographic survey, history inquiry, physical examination, cognitive assessment, blood biochemical test, brain CT/MRI, and gene sequencing. RESULTS: 56.4% (115/204) participants were clinically diagnosed with behavioral variant of FTD, 20.6% (42/204) with nonfluent/agrammatic variant primary progressive aphasia (PPA), 20.1% (41/204) with semantic variant PPA, and 2.9% (6/204) with mixed variant PPA. 11.8% (24/204) subjects harbored the potential causative variants in FTD-related genes, including the MAPT (n = 7), TBK1 (n = 7), GRN (n = 2), TBK1+GRN (n = 1), VCP (n = 1), TARDBP (n = 1), UBQLN2 (n = 1), SQSTM1 (n = 1), DCTN1 (n = 1), HNRNPA1 (n = 1), and C9orf72 GGGGCC repeats (n = 1). The TBK1 T31fs, T457fs, K622fs, c.359-1G>A, the VCP P188T, and the GRN P50fs, P439fs were novel pathogenic/likely pathogenic variants. The TBK1 carriers showed a later disease onset and a higher incidence of parietal atrophy relative to the MAPTcarriers. CONCLUSION: There is genetic and clinical heterogeneity among Chinese FTD population. The TBK1 has a high mutation frequency in Chinese FTD patients.

PSEN2 Mutation Spectrum and Novel Functionally Validated Mutations in Alzheimer's Disease: Data from PUMCH Dementia Cohort.

BACKGROUND: The established causative mutations in the APP, PSEN1, and PSEN2 can explain less than 1%,Alzheimer's disease (AD) patients. Of the identified variants, the PSEN2 mutations are even less common. OBJECTIVE: With the genetic study from the dementia cohort of Peking Union Medical College Hospital (PUMCH), we aim to illustrate the PSEN2 mutation spectrum and novel functionally validated mutations in Chinese AD patients. METHODS: 702 AD participants, aged 30-85, were identified in PUMCH dementia cohort. They all received history inquiry, physical examination, biochemical test, cognitive evaluation, brain CT/MRI, and next-generation DNA sequencing. Functional analysis was achieved by transfection of the HEK293 cells with plasmids harboring the wild-type PSEN2 or candidate mutations. RESULTS: Nine PSEN2 rare variants were found, including two reported (M239T, R62C) and seven novel variants (N141S, I368F, L396I, G117X, I146T, S147N, H220Y). The HEK293 cells transfected with the PSEN2 N141S, M239T, I368F plasmids showed higher Aß42 and Aß42/Aß40 levels relative to the wild-type PSEN2. The PSEN2 L396I, G117X, S147N, H220Y, and R62C did not alter Aß42, Aß40 levels, or Aß42/Aß40 ratio. 1.9%,(13/702) subjects harbored rare PSEN2 variants. 0.4%,(3/702) subjects carried pathogenic/likely pathogenic PSEN2 mutations. The three subjects with the functionally validated PSEN2 mutations were all familial early-onset AD patients. The common symptoms included amnesia and mental symptom. Additionally, the M239T mutation carrier presented with dressing apraxia, visuospatial agraphia, dyscalculia and visual mislocalization. CONCLUSION: The PSEN2 N141S, M239T, and I368F are functionally validated mutations.

Generation of a human induced pluripotent stem cell (iPSC) line from skin fibroblasts of a patient carrying an E363Q mutation in PSEN1 gene.

Alzheimer's disease (AD) is a common neurodegenerative disease characterized by cognitive decline even leading to incapacity, which prevalence estimates that about 5% of AD cases are caused by mutations in genes such as Presenilin-1 (PSEN1). Here we report the generation and characterization of an iPSC line derived from a patient carrying an E363Q mutation in PSEN1 gene. The iPSC line we generated presented a typical morphology, normal karyotype, free from Sendai viral vectors and exogenous factors, expressed endogenous pluripotency marker genes and proteins, which could form embryoid bodies in vitro and form teratoma in vivo as well, demonstrating its pluripotency.

Typical and atypical phenotype and neuroimaging of X-linked adrenoleukodystrophy in a Chinese cohort.

OBJECTIVE: The objective of this study is to describe the typical and atypical clinical and neuroimaging features of ALD in Chinese patients, which will help early diagnosis and intervention to improve prognosis of ALD. METHODS: Forty-one patients in the Leukoencephalopathy Clinic of Neurology Department, Peking Union Medical College Hospital were enrolled. Detailed clinical manifestations and MRI features were analyzed. The relationship between phenotype and genotype as well as biochemical analysis was observed. RESULTS: The patients were classified according to phenotype and onset age, including 14 childhood cerebral ALD (CCALD), 8 adolescent cerebral ALD (adoCALD), 3 adult cerebral ALD (ACALD), 14 adrenomyeloneuropathy (AMN), and 2 ALD in women. AMN was the main presentation in adults. Visual impairment was usual onset symptom in CCALD and cognitive decline and psychiatric symptoms were found in adoCALD and ACALD. Typical MRI feature of CALD was symmetrical peri-ventricular "butterfly wings" like lesions in frontal and/or occipital lobe with peripheral DWI hyperintensities and Gd enhancement. Corpus callosum and internal capsule were always involved. Unilateral lesions were also possible. Cerebral AMN presented with centrum semiovale diffuse involvement. Spinocerebellar variant was a special subtype of AMN with obvious cerebellar and brainstem lesions. No relationships between phenotype and genotype as well as biochemical VLCFAs analysis were found. CONCLUSIONS: We emphasize that corpus callosum and internal capsule are always involved in ALD. A unilateral lesion is also possible. Neuroimaging of cerebral AMN is different from typical CALD with more centrum semiovale involvement. We support spinocerebellar variant was a rare subtype of AMN.

Random Forest Model in the Diagnosis of Dementia Patients with Normal Mini-Mental State Examination Scores.

Background: Mini-Mental State Examination (MMSE) is the most widely used tool in cognitive screening. Some individuals with normal MMSE scores have extensive cognitive impairment. Systematic neuropsychological assessment should be performed in these patients. This study aimed to optimize the systematic neuropsychological test battery (NTB) by machine learning and develop new classification models for distinguishing mild cognitive impairment (MCI) and dementia among individuals with MMSE ≥ 26. Methods: 375 participants with MMSE ≥ 26 were assigned a diagnosis of cognitively unimpaired (CU) (n = 67), MCI (n = 174), or dementia (n = 134). We compared the performance of five machine learning algorithms, including logistic regression, decision tree, SVM, XGBoost, and random forest (RF), in identifying MCI and dementia. Results: RF performed best in identifying MCI and dementia. Six neuropsychological subtests with high-importance features were selected to form a simplified NTB, and the test time was cut in half. The AUC of the RF model was 0.89 for distinguishing MCI from CU, and 0.84 for distinguishing dementia from nondementia. Conclusions: This simplified cognitive assessment model can be useful for the diagnosis of MCI and dementia in patients with normal MMSE. It not only optimizes the content of cognitive evaluation, but also improves diagnosis and reduces missed diagnosis.

Association Between Common Variants of APOE, ABCA7, A2M, BACE1, and Cerebrospinal Fluid Biomarkers in Alzheimer's Disease: Data from the PUMCH Dementia Cohort.

BACKGROUND: The previous studies have identified several genes in relation to Alzheimer's disease (AD), such as ABCA7, CR1, etc. A few studies have explored the association between the common variants, mainly in the non-coding regions of these genes, and cerebrospinal fluid (CSF) biomarkers. Fewer studies target the variants in the coding regions. OBJECTIVE: To illustrate the association between the common variants within or adjacent to the coding regions of AD susceptible genes and CSF biomarkers in AD patients. METHODS: 75 sporadic probable AD patients were extracted from the dementia cohort of Peking Union Medical College Hospital. They all had history inquiry, physical examination, blood test, cognitive assessment, brain MRI, CSF testing of Aß42, 181p-tau, and t-tau, and next-generation DNA sequencing. Sixty-nine common single nucleotide polymorphisms (SNPs) (minor allele frequency > 0.01) within or near the coding region of 13 AD susceptible genes were included in the analysis. RESULTS: The rs7412-CC (APOE) genotype showed lower CSF Aß42 level and higher p-tau/Aß42 ratio than the rs7412-CT genotype. The rs3752246-C (ABCA7) allele correlated with lower CSF Aß42 level. The alternate alleles of six ABCA7 SNPs were related to lower CSF p-tau, including rs3745842, rs3764648, rs3764652, rs4147930, rs4147934 and rs881768. The rs11609582-TT (A2M) genotype showed higher CSF p-tau than the rs11609582-TA genotype. The p-tau/Aß42 ratio was higher in the rs490460-TT (BACE1) genotype relative to the rs490460-GT genotype. CONCLUSION: Some common variants within or near the coding regions of APOE, ABCA7, A2M, and BACE1 are associated with CSF Aß42, p-tau. or p-tau/Aß42.

CSF biomarkers for early-onset Alzheimer's disease in Chinese population from PUMCH dementia cohort.

Introduction: Alzheimer's disease (AD) is one of the highly concerned degenerative disorders in recent decades. Though vast amount of researches has been done in various aspects, early-onset subtype, however, needs more investigation in diagnosis for its atypical manifestations and progression process. Fundamental CSF biomarkers of early-onset AD are explored in PUMCH dementia cohort to depict its laboratory characteristics. Materials and methods: A total of 125 individuals (age of onset <65 years old) from PUMCH dementia cohort were recruited consecutively and classified into AD, non-AD dementia, and control groups. Levels of amyloid-ß 42 (Aß42), total tau (t-tau) and phosphorylated tau (p-tau) were measured using ELISA INNOTEST (Fujirebio, Ghent, Belgium). Students' t-test or non-parametric test are used to evaluate the differences between groups. Area under curve (AUC) of receiver operating characteristic (ROC) curve was introduced to prove the diagnostic powers of corresponding markers. Logistic regression is used to establish diagnostic model to combine several markers together to promote the diagnostic power. Results: The average of all three biomarkers and two calculated ratios (t-tau/Aß42, p-tau/Aß42) were statistically different in the AD group compared with the other two groups (Ps < 0.01). From our data, we were able to provide cutoff values (Aß42 < 570.9 pg/mL; p-tau > 56.49 pg/mL; t-tau > 241.6 pg/mL; t-tau/Aß42 > 0.529; p-tau/Aß42 > 0.0846) with acceptable diagnostic accuracy compared to other studies. Using a combination of biomarkers and logistic regression (area under curve 0.951), we were able to further improve diagnostic efficacy. Discussion: Our study supports the diagnostic usefulness of biomarkers and defined cutoff values to diagnose early-onset AD. We showed that the ratios of t-tau/Aß42 and p-tau/Aß42 are more sensitive than relying on Aß42 levels alone, and that we can further improve diagnostic accuracy by combining biomarkers.

A pilot study of multiple time points and multidomain assessment in cerebrospinal fluid tap test for patients with idiopathic normal pressure hydrocephalus.

OBJECTIVE: Our study aimed to identify the appropriate evaluation time point and assessment forthe CSF tap test(TT) to predict the shunting responsiveness of patients with idiopathic normal-pressure hydrocephalus (iNPH). METHODS: Eighty-eight inpatients with clinically possible iNPH who underwent CSF TT at multiple time points (baseline, 8 hours, 24 hours, and 72 hours after CSF TT) at Peking Union Medical College Hospital were recruited. The multidomain assessment included the timed up and go test(TUG), 10-meter walking tests, and a brief executive function battery. Performance in multidomain assessment at the indicated time points were compared. The positive response rate and cumulative positive rate of multidomain assessment at multiple time points were calculated. And their corresponding specificity and sensitivity of predicting shunting response were calculated according to the follow-up results after shunting. RESULTS: The multidomain assessment performance except TUG at 8 hours were significantly improved at each time point after CSF TT compared with baseline (P<0.01). Reduction more than 10% in the 10-meter walking time and number of steps at 24 hours showed the highest specificity (both 85.7%) and sensitivity (37.5% and 46.7%, respectively) for predicting shunting response. Additionally, an improvement of more than 20% in the composite z score at 72 hours showed 100% specificity and 80% sensitivity for predicting shunting response. CONCLUSION: Multiple time points and multidomain assessment were helpful to identify more shunting responders. Executive function evaluation might be a candidate tool to increase the effectiveness of CSF TT.