Relationship of hypoxia signature with variant subgroup of clear cell renal cell carcinoma (ccRCC) and its association with clinical activity on tivozanib hydrochloride.

Abstract

361 Background: TIVO-1, a randomized Phase III trial in first-line targeted therapy for patients (pts) with ccRCC, demonstrated significant improvement in progression-free survival (PFS) in pts receiving tivozanib hydrochloride (T) vs. sorafenib (S) (11.9 vs. 9.1 months [12.7 vs. 9.1 in treatment-naïve pts]). To further characterize molecular ccRCC subtypes and assess relationships between subtypes and vascular endothelial growth factor tyrosine kinase inhibitor activity, we characterized available molecularly annotated datasets from TIVO-1. Methods: Tumor subtypes were established using hierarchical clustering and evaluated in two microarray ccRCC datasets using gene set enrichment analysis with 51 signatures representing a set of molecular phenotypes. A 9-gene signature comprising genes associated with hypoxia-inducible factor (HIF) transcription was quantified by RT-PCR on all available (69/517) formalin-fixed, paraffin-embedded material from patients using a predefined classifier score and cutoff. Results: Hierachical clustering generated 3 distinct tumor classes. Molecular clusters defining HIF gene expression (low), endothelial cell content (low), extracellular matrix (low), proliferation (high) epithelial cell phenotype (high), and metabolism (high) were differentially expressed in cluster 3 tumors, which represented approximately 15% of the populations. Based on predefined analysis, the hypoxia signature was significantly associated with better PFS on T using a previously established classifier (Table). The hypoxia signature was not significantly associated with PFS on S. There was no significant correlation to single largest diameter for either agent. Conclusions: A novel molecular subtype of ccRCC is characterized by a distinct molecular profile and can be classified by a low hypoxia signature. This hypoxia gene signature may help identify T responders. This signature is seen in subsets of other solid tumors supporting the broad exploration of this candidate T response biomarker. Clinical trial information: NCT01030783. [Table: see text]