DiaGenic has claimed a significant improvement in overall accuracy of a new version of its patented and CE-marked product, ADtect. The data, representing the fourth independent patient study to improve ADtect for blood-based diagnosis of Alzheimer's disease (AD), revealed an accuracy of 82 per cent of AD dementia versus matched control patients.
The higher diagnostic accuracy is obtained from a read-out of 20 genes derived from ADtect, which contains a larger (96) set of genes relevant for the diagnosis of AD. The same selected 20 genes were also tested in two cohorts of patients with amnestic mild cognitive impairment (MCI) that within two years developed AD (MCI/AD) and showed a prediction accuracy for MCI/AD of 70 to 74 per cent versus matched stable MCI (MCI that did not convert to AD). An additional separate 25 non-overlapping gene signature with improved overall prediction was identified predicting conversion from MCI to AD (MCI/AD) with accuracy of 81 per cent.
The results support previous findings that pre-dementia and dementia stages of AD can be identified with a low number of genes in blood evidenced by several independent studies. Conversion from MCI to AD dementia occurs at a rate of about 15 per cent per year and a simple blood test allowing for early diagnosis of AD that is already in the pre-dementia stage is of significant clinical interest.
The increased accuracy of the ADtect test was obtained through an investigation that started with 1,152 genes, selecting only informative genes used in the previously approved ADtect, which includes 96 genes, to a 20-gene set for the improved ADtect. This selection process was done with new bioinformatic algorithms running all patient samples on a qPCR platform from Life Technologies.
Patient samples in the study included 25 patients with dementia in AD, 25 matched controls and 75 patients with MCI. This represents an extension of DiaGenic's previous study cohorts of more than 500 AD and MCI patients.