par Verdoodt, Berlinda;Castelain, Philippe;Bourgain, Claire;Kirsch-Volders, Micheline 
Référence The Histochemical Journal, 27, 1, page (79-88)
Publication Publié, 1995-01
Référence The Histochemical Journal, 27, 1, page (79-88)
Publication Publié, 1995-01
Article révisé par les pairs
| Résumé : | To obtain more information about the relationship between numerical aberrations of chromosome 1 and the overall DNA content of breast cancer cells, fluorescent in situ hybridization with a pericentromeric probe for this chromosome and image analysis based densitometry were carried out on imprints of benign (15 cases, mainly fibroadenomas) and malignant breast disease (31 invasive ductal carcinomas out of 45 cases). The most pronounced aneuploidy was observed in invasive ductal and lobular carcinoma cases both by in situ hybridization and DNA content (76.7 and 75.0% were aneuploid). The frequency of cells with two spots for chromosome 1 was 48.3 and 51.5%, respectively, as compared to 80.3% in control lymphocytes. There was a weak overall correlation (r2 = 0.83) between DNA content and copy number of chromosome 1 in the malignant samples, although some of the DNA diploid/near diploid carcinomas showed a marked aneusomy for this chromosome. Also, some aberrations were present in the benign breast disease samples. Classification of cases by a linear discriminant analysis was most accurate when both techniques were combined (77% of cases correctly classified, according to anatomo-pathological diagnosis). The variables which received the highest weight in the linear discriminant function are the percentage DNA-diploid cells and the fraction of cells with two spots for chromosome 1. The sensitivity and sources of error of both techniques is considered. © 1995 Chapman & Hall. |
