9569 Background: Soft-tissue sarcoma (STS) constitute a heterogeneous group of tumours of mesenchymal origin. Whereas the mainstay of treatment has been surgery and radiation, these tumours are generally considered as quite chemoresistant. However, it is well known that subgroups of patients benefit from chemotherapy. Markers that could predict drug response would therefore be beneficial for the management of this malignancy. We have previously established panel of 17 unique human soft tissue xenografts, representing 7 different histological subgroups and assessed their responsiveness to doxorubicin, ifosfamide, etoposide, and cisplatin. We wanted to utilize these xenografts as a model system to discover for novel candidate marker genes for STS chemo-response. Methods: GE Uniset Human 20K microarrays were used to obtain gene expression profiles from the each xenografts. One-way ANOVA test with a Benjamini-Hochberg multiple test correction allowing a false discovery rate of 5% was used to identify genes with significantly differential expression. Results: Doxorubicin, ifosfamide, etoposide and cisplatin were efficient in 6/17, 10/17, 1/17 and 7/17 xenografts respectively. However, in the expression profiles obtained none of the genes showed significantly correlation with chemo-responsiveness to any of the drugs. Two of the xenografts, TAX 1 and TAX 2, both originate from a malignant fibrous histiocytoma (MFH) in the same patient, but show strikingly different sensitivity to ifosfamide (TAX1 resistant, TAX2 sensitive). When triplicate hybridizations of TAX1 and 2 were compared, 294 genes met the above criteria. In addition we identified a subset of 122 genes that were flagged absent in one of the specimens, present in the other. Among genes with an already described role in mediating drug resistance are GST-pi and glutathione peroxidase. Taken together, these results indicate that discovery of general response markers in STSs may be difficult due to the heterogeneity of the different subgroups constituting this malignancy. Conclusions: Gene expression profiling of the TAX 1 and TAX 2 xenografts revealed a number of interesting candidate marker genes for ifosfamide sensitivity of MFH. This list of genes will be further refined by validation in clinical samples. No significant financial relationships to disclose.
Analysis of Gene Expression Profiles of Soft Tissue Sarcoma Using a Combination of Knowledge-Based Filtering with Integration of Multiple Statistics
