Journal of Clinical Oncology, 2004 ASCO Annual Meeting Proceedings (Post-Meeting Edition).
Vol 22, No 14S (July 15 Supplement), 2004: 502
© 2004 American Society of Clinical Oncology

This Article Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Download to citation manager Rights & Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Hannemann, J. Articles by Van de Vijver, M. Search for Related Content PubMed Articles by Hannemann, J. Articles by Van de Vijver, M.

Abstract

Changes in gene expression profiling due to primary chemotherapy in patients with locally advanced breast cancer

Netherlands Cancer Institute, Amsterdam, Netherlands

502

Introduction: It is currently not possible to predict sensitivity of locally advanced breast cancer (LABC) to specific drugs. This study was designed to identify gene expression patterns in LABC that can predict response to neoadjuvant therapy with Adriamycine (60 mg/m²) and Cyclophosphamide (600 mg/m²) (AC); or Adriamycine (50 mg/m²) and Docetaxel (75 mg/m²) (AD). Methods: We started a phase III trial for patients with locally advanced breast cancer, randomizing between 6 courses of AC or 6 courses AD. Total RNA was isolated from a 14 G core needle biopsy obtained before treatment. All patients underwent surgery after completing chemotherapy. RNA was isolated from any remaining tumor. Amplified mRNA was hybridized on human 18k cDNA microarrays (NKI microarray facility). Supervised and unsupervised classifications were used to analyze differences between gene expression before and after treatment and to correlate expression profiles with response to the chemotherapy. Results: 62 patients with LABC have been randomized in the study. Good quality RNA from tissue with more than 50% tumor cells was obtained from 46 biopsies and from 18 tumors after chemotherapy. Data from 49 patients were included in the analysis: 25 from the AC arm, 24 from the AD arm. Using two-dimensional unsupervised hierarchical clustering, the tumor and biopsy clustered together in almost all cases when no response to chemotherapy was observed. When partial or complete remission was observed, the untreated biopsy specimens and the tumors after treatment clustered together more closely than the specimen obtained from individual patients before and after treatment. Using 9-fold cross-validation on the whole dataset containing 46 biopsies before and 18 specimen after treatment, an optimal number of 30 genes could be identified to distinguish between specimen before and after treatment with an average performance of 91%. Discussion: When breast carcinomas do not respond to neoadjuvant chemotherapy, there are no large alterations in the gene expression pattern. When tumors respond, a gene expression pattern of 30 genes distinguishes tumors before and after therapy.

No significant financial relationships to disclose.

Abstract presentation from the 2004 ASCO Annual Meeting