Abstract Details

Title Can Transcriptomics Add to the Diagnosis of Multiple Sclerosis?

Topic MS and Related Diseases

Presentation(s) P03 - (-)

Poster/Presentation
Number 227

Objective

Background BACKGROUND: In the absence of pathognomonic clinical features or laboratory test, the diagnosis of MS is based on clinical symptomatology and the presence of central nervous system lesions disseminated in time and space with no better explanation. Early accurate diagnosis is important for appropriate therapy.

Design/Methods DESIGN/METHODS: Microarray gene-expression analysis (Affimentrix U133A-2 microarrays) was performed on blood samples obtained from 79 relapsing- remitting MS (RRMS) patients (47 females, age 33.3±0.9 years, EDSS 1.7±0.2, disease duration (DD) 3.2±0.4 years) and 31 patients with MS-resembling symptoms and white matter brain lesions that fail to fulfill MS diagnostic criteria (NonMS, 23 females, age 39.4±1.2 years, DD 2.0±0.5 years). A signature of differentially expressed genes between MS and NonMS groups (p<0.05 by false discovery rate correction) was established. Then, cross- validation routine based on Least-square and K-neighbor algorithms was used for gene-expression based classifiers generation. Potential classifiers were validated on additional independent blood samples from 28 RRMS patients (14 females, age 34.8±1.5 years, EDSS 1.8±0.21, DD 2.9±0.7 years) and 14 NonMS patients (11 females, age 38.0±1.6 years, DD 5.8±1.7 years).

Results RESULTS: A total 540 genes differentiated between MS and NonMS patients. This signature was enriched by over-expressed genes associated with Neurotrophin/TRK Signaling (p=1.4*10-4), T-cell related development (p=1.5*10-4), differentiation (1.6*10-3), and proliferation (p=5.8*10-6). A-10-gene-classifier demonstrated correct classification rate of 88.2±3.0% with sensitivity 91.0±3.2% and specificity 80.0±7.1% between MS and NonMS groups. In the independent validation group of 42 patients, the classifier demonstrated correct classification rate of 76.2±6.6 % with sensitivity of 79.0±7.6% and specificity of 0.71±12.0%.

Conclusions CONCLUSIONS: Blood gene-expression profiling is a powerful tool, that could be applied in clinical decision making for MS differential diagnosis.

Authors/Disclosures
Michael Gurevich (Sheba Medical Center) PRESENTER	No disclosure on file
Gadi Miron	No disclosure on file
David Magalashvili, MD	No disclosure on file
Mark Dolev Dolgopiat, MD	No disclosure on file
Anat Achiron, MD, PhD, FAAN (Sheba Medical Center, Tel-Hashomer)	Dr. Achiron has nothing to disclose.
	No disclosure on file

��ɫ��

��ɫ��