• Citer

Digital cell image analysis of Ewing's sarcoma.

par Kiss, Robert ;Larsimont, Denis ;Huvos, A G
Référence Analytical and quantitative cytology and histology, 13, 5, page (356-362)
Publication Publié, 1991-10

Article révisé par les pairs

• Détails
• Contenu

Résumé :

We analyzed the lesional tissue in Ewing's sarcoma by means of a cell image processor computing 15 nuclear parameters in order to quantify the morphologic variability of the tumor cell nuclei. To this end we combined 32 cases (350-400 Feulgen-stained nuclei analyzed per case) in a data file, which was then subjected to principal component and canonical analyses. We found considerable heterogeneity within the cell nucleus population of Ewing's sarcomas. Indeed, after the arbitrary subdivision of the file into two cell classifiers (CC1 and CC2 cell types) on the basis of the first canonical function, the 32 Ewing's sarcomas showed a great deal of variability in the proportion of CC1 and CC2 cell nucleus types. The nuclei of the CC1 type had a more finely textured chromatin when compared to the CC2 type, the nuclei of which exhibited a more granular chromatin pattern. Additionally, these 32 Ewing's sarcomas were characterized by three distinct DNA histogram types. Eight tumors displayed a diploid nonproliferating DNA histogram pattern (type A), 11 a diploid proliferating (type B) and 13 an aneuploid (type C) DNA histogram profile. We found a highly significant relationship between these DNA histogram types and the CC1:CC2 cell type percentage ratio. The eight Ewing's sarcomas with a type A DNA histogram contained a significantly higher proportion of CC1 cell type nuclei as compared to the 13 tumors with a type C DNA histogram, which contained a great proportion of CC2 cell type nuclei.