Autopsy-Confirmed Minimally Invasive Biomarker Identifies Alzheimer's Disease Even in the Presence of Co-Morbid Pathologies

Autopsy-Confirmed Minimally Invasive Biomarker Identifies Alzheimer's Disease even in the Presence of Co-Morbid Pathologies

Background:

Over the past two decades, clinical trials that tested potential Alzheimer’s disease (AD) therapeutics typically included patients without a definitive neuropathological diagnosis for AD dementia. An early diagnosis of AD can help determine if the symptoms a patient is experiencing are genuinely due to AD or some other conditions that can be curable. Moreover, clinical diagnoses are often uncertain, particularly during the first 4 to 5 years of disease duration. In more than 50% of cases, AD pathology remains co-morbid with pathologies of other non-AD neurodegenerative diseases, e.g., Vascular dementia, multi-infarct dementia, Pick’s disease, Lewy-body dementia, Parkinson’s disease, Frontotemporal dementia, tauopathy, corticobasal degeneration, etc.  This urgent unmet medical need for a highly accurate, easily accessible AD biomarker has motivated the development of several candidate AD tests. These have included MRI and PET imaging of amyloid plaques, CSF and plasma measures of soluble amyloid and tau, and blood levels of tau. While some of these methods have promising sensitivity, none have achieved comparably high specificity when referenced to autopsy-pathologic findings of AD vs. non-AD dementia (non-ADD). The CLIA- certified DISCERN test, on the other hand, has shown high ( > 95 %) sensitivity and high ( > 95 %) specificity with autopsy validation, and was derived from several factors including synaptic loss, neuronal death, inflammation, amyloid deposition, and hyperphosphorylation of tau protein. There are currently no FDA-approved tests that are highly sensitive and highly specific for AD. Diagnostic accuracy has been even more difficult due to recent findings that >50% of AD brains show co-morbid pathologies such as multi-infarct dementia and frontal lobe dementia. The data below describes an autopsy-confirmed AD Biomarker that identifies AD pathology even with co-morbid brain pathology.

Purpose/Objectives:

A Morphometric Imaging (MI) assay was previously shown to correlate dementia and the presence of AD pathology in the brains of AD patients (Chirila et al., 2013) with abnormalities of skin fibroblasts isolated with routine punch biopsies (~3mm). After the skin biopsy, pure fibroblast culture was established for each patient. Cells were cultured on a thick layer of 3-D Matrigel matrix and subjected to image analysis. AD cell lines formed large aggregates in contrast to non-AD dementia (non-ADD) or non-demented control (NDC) cell samples. Typically skin fibroblasts formed “networks” analogous to networks formed by neurons isolated in cell culture and AD networks were slower and less connected Quantitative image analyses enabled the calculation of average unit aggregate area (A) in terms of ln(A/N). Samples were collected with a double-blind protocol for demented patients > 55 years old who eventually reached blinded autopsy examination.

Methods:

In general, AD cell lines formed a smaller number of larger aggregates than non-AD cell lines. Such a difference enabled counting the number of aggregates (N) and measuring aggregates’ average area (A). The quantitative measure, Ln(A/ N) at 48 hours, is plotted for AD and non-ADD patients from the validation cohorts. This method accurately diagnosed AD patients and distinguished them from non-ADD patients. The probability distributions of the morphometric imaging signals showed a clear separation of the AD sample measurements (n= 25) from the non-ADD group values (n= 21). AD cell lines formed fewer large aggregates (cells clustered together) than non-AD or NDC cell lines. The cut-off value of Ln(A/N) =6.98 was determined from the biomarker values of NDC cases.

Based on these results, the sensitivity and specificity of the MI assay for diagnosing AD were: Sensitivity= 100 % (95 % confidence interval 86 - 100 %); Specificity= 100 % (95 % confidence interval 84 - 100 %). For the Biomarker AD vs. Non-ADD data, True Positive= 26, False Negative= 0, and False Positive= 0, with Sensitivity and Specificity calculated as 100 and 100, respectively. (AD vs. non-ADD, p < . 000001). This AD specificity was observed even in the cases with pathologic co-morbidity (i.e., " Mixed Dementia”) that included AD with pathologies of other dementias such as Parkinson’s disease, Pick’s disease and Frontal Lobe Dementia.

Results:

The CLIA-certified MI assay measures the disease effects in living cells, and its behavior is a way that measures a whole ensemble of defects (e.g., mitochondrial dysfunction, imbalance in kinases/proteases, dysfunction in cellular energy metabolism, cellular senescence, Ca2+ homeostasis, dysfunction in proteasome activity (removal of toxic proteins, e.g., tau, A-beta, etc., and others). In these autopsy-confirmed results, the MI Biomarker distinguished AD from non-ADD patients and correctly diagnosed AD even in the presence of other co-morbid pathologies at autopsy. Therefore, this highly accurate, minimally invasive AD biomarker was validated by rigorous reference to the NIH gold standard criteria for an AD diagnosis, dementia in life, and the presence of plaques and hyperphosphorylated tau at autopsy. MI assay results should be invaluable for AD patient stratification for AD therapeutic clinical trials. 

Authors report ownership interest in NeuroDiagnostics LLC dba SDx.  

Conclusions:

References: