Cell Biochemistry Martinsried

Predictive Medicine by Cytomics

from: Purdue CD-ROM Vol.6   2002   ISBN: 0-97117498-3-3

1. Problem:
Diseases are caused by molecular changes in cellular systems or organs. They are induced by exposure to external influences like microorganisms, allergens, toxic substances a.o. or alternatively by genetic disposition or genetic aberrations. Disease course prediction ( >95% correct) for individual patients is usually considered impossible for the majority of diseases. Exceptions concern e.g. genotypic aberrations detected during amniocentesis or preimplantation diagnostics (PID).
The substantial clinical interest in predicting the future disease development in individual patients prompts for the search of the disease relevant molecular information at the cellular level.
Considering genomics or proteomics for genome or proteome analysis, the high multiparametric complexity and the usual preanalytic content mixing from different cell populations in cell or tissue extracts, constitute substantial drawbacks for predictive conclusions in common diseases like infections, allergies, malignancies, intoxications, degenerative disease a.o.

2. Predictive Medicine by Cytomics
Cytomics, the multimolecular cytometric analysis of the cellular heterogeneity of cytomes (cellular systems/organs/body), access a maximum of information on the apparent molecular cell phenotype as it results from cell genotype and exposure.

The cell phenotypes in the naturally existing cellular and cell population heterogeneity of disease affected body cytomes contain the information on the future development (prediction) as well as on the present status (diagnosis) of a disease.

Data classifications are considered predictive for individual patients at predictive values >95% for each classified disease category of the learning set while they are prognostic at values <95%.

General concept for predictive medicine by cytomics:
a.) multiparametric cytometric determination of functions or constituents in disease associated cytomes
b.) exhaustive analysis (1, 2) of all measured numeric parameters for all cell populations (i.e. in practice for >95% of the collected cells)
c.) data pattern classification of this entire information against the future disease course of patients during the learning phase
d.) classification of the embedded test set of patient data, measured under the same conditions as the learning set but remaining unknown to the learning process. Typically, every 5th or 10th patient is assigned to the test set prior to the learning phase to exclude classification biases.
e.) prospective classification of data collected from subsequent new patients during the clinical evaluation phase

3. Data Collection
Patterns of various biomolecules can be reliably quantitated by cytometric analysis of viable or fixed cells following staining with biochemically specific fluorescent dyes. The particular effort of this laboratory consists in the development of specific stains for cell functions in viable cells as sensitive indicators of the altered cellular metabolism in acute or chronic disease. The simultaneous multiparameter data collection by the cytometer provides high amounts of functional and structural information on heterogeneous i.e. essentially unprocessed ex-vivo cell suspensions shortly after removal from the human body.
The cellular heterogeneity of human samples offers important advantages for clinical and experimental system cytometry (cytomics) because the high information content of simultaneously collected multiparameter data from a great variety of different cell types can be utilized. The cytometric strategy is explicitely to measure as much heterogeneity as possible to profit during evaluation from the high information content of biocomplexity. The cytometric approach is therefore quite different from the tissue biochemistry approach where one tries to reach as much homogeneity as possible e.g. by the isolation of groups of similar cells by laser microdissection to assure unambiguous interpretation of experimental results.

4. Multiparameter Data Pattern Classification
The exhaustive extraction of information from cytometric or clinical chemistry multiparameter measurements by a laboratory and instrument independent, self learning and standardized data classification algorithm, developed earlier (2), provides accurate single patient disease course prediction, as well as diagnostics in case of sufficiently information rich molecular data or other patient data.

5. Examples
Clinical examples from several medical disciplines underline this point. Predictive medicine by cytomics represents Evidence Based Medicine (EBM) at the cellular level.
The practical consequence of this approach is that complications in a number of common diseases like severe infections, shock, exacerbation of rheumatoid and asthmatic disease, thromboembolic complications in diabetes, myocardial infarction and stroke sensitive patients or survival in cancer patients, but also e.g. complications in bone marrow stem cell transplantation (BM-SCT) will become increasingly predictable at the individual patient level.
Minor interventions like cytometry supervised short term antiphlogistic therapy e.g. just prior to an imminent exacerbation of rheumatoid disease may prevent severe tissue destruction leading otherwise to the stepwise disabling of the patient by deficient repair processes. The cell biochemical approach has in this case the potential to significantly postpone the invalidization of patients. The higher quality of patients's life would be paralleled by shorter disease periods at substantially lower therapy costs and a lower number of unwanted therapeutic side effects (Optimized Medicine).

© 2024 G.Valet
1965-2006: Max-Planck-Institut für Biochemie, Am Klopferspitz 18a, D-82152 Martinsried, Germany
Last Update: Mar.25,2002