par Pauwels, Olivier ;Kiss, Robert
Référence Journal of cancer research and clinical oncology, 119, 9, page (533-540)
Publication Publié, 1993
Article révisé par les pairs
Résumé : We illustrate the potential application of digital cell-image analysis to characterize the morphonuclear modifications induced by various drugs including VP16, a podophyllotoxin derivative, PE1001, an investigational alkylating agent, and doxorubicin, an intercalating agent. Fifteen parameters representative of morphometric (nuclear area), densitometric (nuclear DNA content), and textural (chromatin organization, condensation, and distribution) features were computed on Feulgen-stained nuclei obtained from fine-needle aspirations serially performed during treatment on the MXT mouse mammary cancer model. We observed marked differences between the control and drug-treated MXT cell nuclei. However, mathematical data processing was necessary to improve the ratio of the chemotherapy-induced morphonuclear signal to the control biological morphonuclear signal. This data processing relies upon the use of principal-component analysis followed by the canonical transformation of data. The present method can be applied to all human cancers on which fine-needle aspiration can be performed.