Résumé
Abstract Activation of the ERBB2 oncogene seems to be an early event in breast cancer progression and prevalent in in situ carcinomas. However, its prognosis value, albeit recognized for node-positive patients, remains controversial for those without apparent
nodal involvement. One possible reason for this problem is likely to be the difficulty of defining threshold levels for ERBB2 protein overexpression. ERBB2 protein expression was therefore analyzed in primary invasive breast tumors. Quantification of the gene product by a commercial ELISA
test was compared to results obtained by immunohistochemistry and western blotting, as well as to gene amplification status determined by Southern blotting. Correlations between results obtained by the different techniques were highly significant (P value < 10−6).
Nevertheless, ELISA permitted us to determine three levels of protein expression corresponding to distinct tumor subsets. 1) Tumors with pl85/ERBB2 expression levels exceeding 10 U/μg exhibited in most cases amplification of the gene (83% of cases), DNA aneuploidy (81%) and absence of estrogen
receptor (ER) (44%). Such high levels of protein expression were exclusively observed in invasive ductal carcinomas and were prevalent in those showing a significant in situ component. 2) "Intermediate" levels of expression (3-10 U/μg) were rarely observed in tumors
exhibiting gene amplification (9%), but were preferentially found in cancers of more favorable prognosis (only 49% were aneuploid and 9% estrogen receptor negative). 3) Levels of pl85/ERBB2 below 3 U/μg were detected in benign mastopathies and, thus, carcinomas presenting such levels were
scored ERBB2 negative. Interestingly, invasive lobular carcinomas were rarely ERBB2 positive, and if so, only at intermediate levels. Moreover, our data show a complex interrelationship between pl85/ERBB2 expression and ER levels. Indeed, tumors with more than 10 U/μg of p185 were prevalently
ER, whereas those with pi85 ranging from 3 to 10 U presented elevated levels of ER.