1. ¹Department of Pathology, Centre Hospitalo-Universitaire Nord, Marseille, France
2. ²Department of Biopathology, Institut Paoli Calmettes, Marseille, France
3. ³UMR 599 INSERM, Institut Paoli Calmettes, Marseille, France
4. ⁴Faculté de Médecine Timone, Marseille, France
5. ⁵Department of Gynecology and Breast Oncology, Centre Hospitalier Universitaire de la Conception, Centre Hospitalier Privé Beauregard, Marseille, France

Correspondence: Professor C Charpin-Taranger, E-mail: colette.charpin@ap-hm.fr

Received 13 July 2006; Revised 23 October 2006; Accepted 29 November 2006.

Abstract

Inflammatory breast carcinoma (IBC) is a rare but aggressive tumour associated with poor outcome owing to early metastases. Increased expression of c-Met protein correlates with reduced survival and high metastatic risk in human cancers including breast carcinomas and is targetable by specific drugs, that could potentially improve the prognosis. In the present study, we compared c-Met expression in IBC (n=41) and non-IBC (n=480) immunohistochemically (Ventana Benchmark autostainer) in two tissue microarrays (TMA) along with PI3K and E-cadherin. The results were quantified through an automated image analysis device (SAMBA Technologies). We observed that (i) c-Met was significantly overexpressed in IBC as compared with non-IBC (P<0.001), (ii) PI3K was overexpressed (P<0.001) in IBC, suggesting that the overexpressed c-Met is functionally active at least through the PI3K signal transduction pathway; and (iii) E-cadherin was paradoxically also overexpressed in IBC. We concluded that overexpressed c-Met in IBC constitutes a potential target for specific therapy for the management of patients with poor-outcome tumours such as IBC. Automated image analysis of TMA proved to be a valuable tool for high-throughput immunohistochemical quantification of the expression of intratumorous protein markers.