Large discrepancy remains for the hormone-responsible cancers with regards to the conditions generating the optimal opportunity for cancerogenesis. In the research, altered steroid profiles were observed in patients with urogenital tract cancer diseases, namely bladder, kidney, prostate and testis ones. The presented steroid profiles from 154 subjects, including 77 urogenital tract cancer patient and 77 healthy controls were determined by liquid chromatography coupled to mass spectrometry method. Because the original experimental data obtained as a result of analytical experiment in order to interpret them in better way required the appropriate pre-treatment, the data were standardized by scaling and centering. In order to determine which samples form a collection for a high-capacity predictive model, Kennard-Stone's algorithm was used. A principal component analysis of preprocessed data provided better consistency of the steroid profiles with health status of subjects than PCA profiles without data preprocessing and showed a tendency to separate clusters of cancer patients from healthy subjects. The discriminant analysis was also performed and the percent of correct classification of cancer patients and control group was calculated. Finally, detailed studies examined the role of steroid profiles measured in urine, and considered as potential biomarkers related to urogenital cancer and associated renal dysfunctions.
Copyright © 2012 Elsevier B.V. All rights reserved.