Journal of Clinical Oncology, 2004 ASCO Annual Meeting Proceedings (Post-Meeting Edition).
Vol 22, No 14S (July 15 Supplement), 2004: 500
© 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 Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Yoshimoto, M. Articles by Noda, T. Search for Related Content PubMed Articles by Yoshimoto, M. Articles by Noda, T.

Abstract

Prediction of the therapeutic response to paclitaxel by gene expression profiling in neoadjuvant chemotherapy for breast cancer

Cancer Institute Hospital, Tokyo, Japan

500

Background: Paclitaxel is one of the most active agents for breast cancer, but the sensitivity is heterogeneous. In order to avoid unnecessary treatment, identification of a reliable predictive marker is desired to distinguish between patients who are likely to respond and those who are not. We report the discovery of a gene expression profile that predicts response to paclitaxel in breast cancer patients. Methods: We took core needle samples from 75 patients with primary breast cancer (size > 3cm) before treatment and then assessed tumor response to neoadjuvant weekly paclitaxel (80 mg/m² x12 cycles) under IC. Clinical response rate was 75%, and pCR rate was 3%. Patents were divided into five groups according to clinical and pathological responses (Group-1, extremely resistant; Grp-2, resistant; Grp-3, moderate responder; Grp-4, responder; Grp-5, high responder). RNA extracted from biopsy samples using microdissection method were profiled on cDNA microarrays of 23,000 human transcripts. Differentially expressed 197 genes between high responder (n=7) and extremely resistant (n=7) groups were selected by Mann-Whittney U-test (p<0.05). Secondly, machine-learning method (AdaBoost) was performed to determine the greatest estimated accuracy between 24 responders (Grp-3,4,5) and 16 non-responders (Grp-1,2), and high-scored predictive set of 23 genes were selected. At the time of this submission, tests were performed on 40 pts. Results: In leave-one-out cross validation analysis using the 23 genes, all responders and non-responders were correctly classified with an accuracy of 100%. Correlation between RNA expression measured by the arrays and semiquantitative RT-PCR was also ascertained. Conclusions: If validated, this gene expression profile could allow development of a clinical test for paclitaxel sensitivity, and may help physicians to select individual patients who are likely to benefit from paclitaxel.

No significant financial relationships to disclose.

Abstract presentation from the 2004 ASCO Annual Meeting