Abstract
I) Amyloid Beta Brain PET Imaging and Blood Biomarkers for validation study for diagnosis of Alzheimer disease dementia Objectives: To compare the quantitative data of Amyloid Beta (Aβ) PET neuroimaging with blood-based biomarkers in AD, non-AD dementia, and elderly control, identify suitable cut-off values in the differential diagnosis of these subgroups, and correlate with visually positive Aβ- PET. Materials and Methods: We conducted a prospective study in 25 patients with clinically diagnosed dementia (16 AD and 9 non-AD) and 5 cognitively normal elderly controls. Aβ-PET using 18F-florbetapir and plasma biomarkers (Aβ40, Aβ42, and p-tau181) were obtained in all subjects within 6 months before or after the PET study. CSF was also performed in 11 dementia patients. A nuclear medicine physician visually interpreted 18F-florbetapir PET images. The quantitative analysis of Aβ-PET to obtain Centiloid (CL) followed the standard method using the SPM8 pipeline. A P-value of < 0.05 is considered statistically significant. Conclusion: Aβ-PET imaging and plasma biomarkers show promising diagnostic performances and potential clinical usefulness in diagnosing dementia. Plasma p-tau181 best differentiates AD from normal control, while Aβ-PET CL best differentiates AD from non-AD dementia, followed by Aβ42/40 ratio, Aβ42, and Aβ40. Using the CL cut-off >10.87, the sensitivity and specificity to identify visually positive Aβ-PET reaches 100%, with varied cut-off values and performance levels from plasma biomarkers. II) The association between blood biomarkers (Aβ40, Aβ42, p-tau, NFL by Singlemolecule array (SIMOA) essay) and cerebrospinal fluid (CSF) biomarkers for dementia (CSF Aβ42, p-tau, t-tau by ELISA method) Objectives: To study the association of between CSF biomarkers of Alzheimer disease (AD) and plasma biomarkers of AD. To study the association of between plasma biomarkers of AD, cognition, function, and other blood biochemistry parameters. Materials and Methods: 74 subjects (45 with AD, 29 with non-AD) had plasma biomarkers, CSF biomarkers of AD, blood biochemistry study, cognitive evaluation and activity of daily living function evaluation were recruited. Spearman correlation analysis was performed. P<0.05 was regarded as statistical significance. Results: CSF Aβ42 had low correlation with plasma Aβ42:40, had low negative correlation with plasma pTau181, but negative medium correlation with plasma GFAP. CSF pTau181 showed medium correlation with plasma pTau181, negative medium correlation with plasma Aβ42:40, and negative low correlation with plasma Aβ42. CSF total Tau showed negative low correlation with plasma Aβ42:40 but medium correlation with plasma GFAP, and plasma pTau181. All plasma biomarkers of AD were influenced by age, education, body mass index (BMI), global cognition, and daily function. Levels of plasma NFL were associated with blood BUN, creatinine, cholesterol, triglyceride, LDL, albumin, total protein, and AST liver function. Conclusion: Low correlation between CSF and plasma biomarkers for AD had been confirmed in our study. Cognition and daily function were related to the change in biomarker levels. Systemic parameters were related to plasma NFL. Therefore, utilization and interpretation of plasma NFL for diagnosis of AD should be cautious. III) Correlation between blood biomarkers and neuropsychiatric symptoms in dementia and mild cognitive impairment Thai cohort Objectives: We explored the relationship of neuropsychiatric symptoms (assesses by Neuropsychiatric Inventory, NPI) and other Alzheimer disease pathophysiology from blood Aβ42:40, GFAP, NFL, and pTau181. Materials and method: 222 subjects included 96 dementia, 66 MCI, and 60 normal controls (NC). 56.25% of dementia group were AD and atypical AD. 62.12% of MCI were amnestic type. Spearman correlation was used to calculation the correlation coefficient between blood biomarkers and NPI subscale scores. T test and Mann-Whitney U test were used to compare biomarkers of those with and without subitem symptoms of NPI to test whether two sample mean or median scores are equal or not. Factor analysis of subscale NPI severity score was analyzed. Regression analysis was calculated the association between the presence of NP symptoms and blood biomarkers of dementia. P<0.05 was regarded as statistical significance. Results: Mean ages were 71.26±9.82 years old in dementia group, 68.83±8.01 years old in MCI group, and 59.28±9.22 years old in NC group. The sum NPI subscale scores showed significant low correlation (r=-0.193-0.294) were correlated with blood Aβ42:Aβ40, Ptau181, NFL, and GFAP.
The presence of NPI symptoms demonstrated significant low correlation with age (r=0.142) and global cognition (assessed by Thai mental state examination, TMSE, r=-0.337). Blood pTau181 displayed significant low correlation with SEVERITY scores of hallucination, depression, anxiety, apathy, and irritability. There was a significant low correlation between blood Aβ42:Aβ40 and NPI severity scale scores of delusion, anxiety and irritability. Our factor analysis of NPI severity scores showed 5 factors namely psychotic factor, frontal factor, mood factor, anxiety, and hyperactivity with total variance of 61.35%. Blood NFL showed significant low correlation with most of frontal behaviors (euphoria, apathy, disinhibition, irritability, aberrant motor activity) and hallucinations. Inflammation process, assessed by blood GFAP, appeared to be significantly related with psychosis, and frontal behaviors. After controlled by age; the presence of NP symptoms still had low correlation with blood pTau181 and GFAP. Regression analysis revealed that higher blood levels of pTau181, NFL, and GFAP were associated with significant odds ratio (ORs) of the presence of NP symptoms. Lower blood levels of Aβ42:Aβ40 were associated with significant ORs of the presence of NP symptoms. Higher blood pTau181 was associated with significant ORs of the presence of hallucinations, apathy, and irritability. Lower blood Aβ42:Aβ40 was associated with significant ORs of delusions and irritability. Higher blood NFL was associated with significant ORs of the presence of apathy; and higher blood GFAP was associated with significant ORs of the presence of hallucinations, anxiety, euphoria, apathy, disinhibition, and irritability. Conclusion: We discovered that psychosis and mood factors of NP symptoms were related to blood pTau181. While, most frontal factors of NP symptoms showed significant relationship with axonal destruction and microglia activation/inflammation assessed by blood NFL and GFAP. However, the presence of irritability symptoms had significant differed levels of blood pTau18, Aβ42:Aβ40, NFL and GFAP. These results can support the strength of blood Aβ42:Aβ40, pTau181 and GFAP biomarkers for predicting behavioral symptoms in cognitive impairment individuals. IV) Cost-utility analysis of blood-biomarker tests for Alzheimer’s disease diagnostic process and predict progression from mild cognitive impairment to dementia: a modelling study OBJECTIVE: To assess the cost-utility of employing Single-molecule array (Simoa) technology to measure blood biomarkers, specifically Aβ42:40 protein, in the diagnostic testing of Alzheimer's disease (AD) and mild cognitive impairment (MCI due to AD) among elderly individuals at high risk of dementia. METHODS: A decision tree, integrated with Markov models, was constructed from a societal perspective. The analysis used a 1-year cycle length to estimate costs and health outcomes over a lifetime horizon, applying a 3% annual discount rate. The study compared the diagnostic approach using Simoa technology for Aβ42:40 protein measurement (Simoa strategy) with the current scenario of no testing for elderly individuals at high risk of dementia (no testing strategy). Sensitivity and specificity data for Simoa were primarily evaluated in Thai patients in Siriraj hospital. Non-pharmacological treatment efficacy was sourced from the FINGER study. Epidemiological data, transition probabilities, and direct medical costs related to AD were obtained from literature reviews. Direct non-medical costs and utility values were primarily obtained from Siriraj Hospital. Lifetime cost, quality-adjusted life years (QALYs), and the incremental cost-effectiveness ratio (ICER) were calculated and assessed against the costeffectiveness threshold of 160,000 THB per QALY gained.Probabilistic and one-way sensitivity analyses were performed to estimate parameter uncertainties. RESULTS: Using Simoa technology to detect Aβ42:40 protein for AD and MCI due to AD diagnosis provided an additional 0.010 life-years (LY) gained and 0.017 QALYs gained, incurring additional costs of 16,280 THB compared to the no testing strategy. The ICER for Simoa was 944,487 THB per QALY. The results were sensitive to the efficacy of non-pharmacological treatment. CONCLUSIONS: Implementing Simoa technology for Aβ42:40 protein detection in the diagnosis of AD and MCI due to AD was found to be not cost-effective in Thailand. Lowering Simoa technology's price by 83% could potentially enhance its cost-effectiveness, emphasizing the importance of price negotiation for ensuring financial value and affordability.
