Description:
<jats:p> 611 </jats:p><jats:p> Background: Improved understanding of the pathobiology of the metastatic cascade as well as the identification of new prognostic markers may lead the path to the development of novel targeted agents in breast cancer patients (BC pts). Recently, HER3-expression was postulated as independent risk factor for metastatic spread. Methods: Pts of different BC subtypes (luminal, HER2-amplified, triple-negative) with metastatic disease were identified from a breast cancer data base. Tissue of the primary tumor was retrieved from the local pathology institute. Immunohistochemical staining of estrogen-receptor, progesterone-receptor, and HER2 and HER3 was performed. In HER2 equivocal cases, subsequent FISH analysis was performed. Results: Specimens of 110 pts (36/110 luminal, 35/110 HER2-amplified, 40/110 triple-negative) were available for this analysis. 23/110 (21%) specimens showed strong, complete, membranous staining for HER3 of at least 10% of all tumor cells. HER2/HER3 co-expression was observed in 12/110 (11%) specimens. HER3 showed a statistically significant association with HER2-expression (p=0.02; Chi square test). No correlation was observed for HER3-expression and overall survival (OS), incidence of brain metastases, or time to diagnosis of brain metastases in the entire patient cohort (p>0.05; log rank). In the HER2-amplified subgroup, however, HER3-expression was significantly associated with shorter OS (median 30 vs. 63 months; p=0.02; log rank test) and remained significant when entered into a multivariate model (p=0.02; Cox regression). Conclusions: HER2/HER3 co-expression is significantly associated with impaired OS in pts with HER2-positive metastatic breast cancer. Co-inhibition of HER2 and HER3 or inhibition of HER2/HER3 hetero-dimerization could improve prognosis of this patient population. </jats:p>