Cytomics and predictive medicine for oncology

Abstract

Cytomics combines the multimolecular cytometric analysis of cell and cell system (cytome, cytomes) heterogeneity on a single cell level with the exhaustive bioinformatic knowledge extraction from all analysis results (cytomics = system cytometry + bioinformatics). It therefore yields a maximum of information about the apparent molecular cell phenotype. At present, in the typical hypothesis driven way the high amount of information collected by multiparameter single cell flow-or slide-based cytometry measurements is preferentially used to investigate the molecular behaviour of specific cell populations in the perspective of the hypothesis. The information outside the scope of the hypothesis remains frequently unused. In contrast, under the predictive medicine by cytomics concept, the entire available information is processed (sieved) in a data driven way under the general data mining hypothesis that such data may contain useful information for clinical diagnosis and especially for therapy related predictions about disease progress in individual patients. The present experience from clinical data sets of various malignant and other diseases suggests that this is a promising concept for cancer patients since it has amongst others the potential to identify high risk patients prior to an anticipated therapy as being unsusceptible with accuracies of greater 95% or 99%. This opens the way for early decision on alternative therapies by objective and molecularly standardised criteria. This has been traditionally difficult by current prognosis evaluation according to the widely used Kaplan-Meier statistics for patient groups. The cytomics concept is also useful for cancer research in general because it favours the enrichment of informative parameters concerning disease outcome in individual organisms or cell cultures from an essentially unlimited number of parameters. The selected parameters are useful as a starting point for mathematical modelling in systems biology without requirement for detailed pre-existing knowledge about potential disease inducing mechanisms. It has therefore the potential for the discovery of new molecular cell pathways and for their subsequent molecular reverse engineering. © 2010 Springer Science+Business Media B.V.