Phase II study of the HSP90-inhibitor BIIB021 in gastrointestinal stromal tumors.

BACKGROUND: HSP90 inhibition leads to proteosomal degradation of activated KIT and has in vitro activity against gastrointestinal stromal tumors (GIST). BIIB021 is an oral non-ansamycin HSP90 inhibitor. We carried out a phase II study of BIIB021 in patients with GIST refractory to imatinib and sunitinib. PATIENTS AND METHODS: The primary end-point was metabolic partial response (mPR) as assessed by fluorodeoxyglucose positron emission tomography (FDG-PET). The secondary end-points were pharmacokinetic assessments of BIIB021 and pharmacodynamic assessments of HSP70. Twenty-three patients were treated on two schedules: 12 pts received 600 mg twice a week (BIW) and 11 patients received 400 mg three times a week (TIW). All had prior imatinib and sunitinib but stopped>14 days before starting BIIB021. RESULTS: The median age was 59 years (33-88 years), 61% male, 44% Eastern Cooperative Oncology Group 1 (ECOG1). The best response was PR by FDG-PET for five patients (3/12 at 600 mg BIW and 2/9 at 400 TIW) for an overall response rate of 22%. The response duration was 25-138 days. Adverse events (AEs) were mild to moderate. The mean Cmax was 1.5 µmol and the mean AUC was 2.9 µmol h. Cmax>1.5 µmol was associated with a decrease in standardized uptake value (SUVmax). HSP70 increased substantially following treatment. CONCLUSIONS: This study met its primary end-point. BIIB021 leads to objective responses in refractory GIST patients. Pharmacodynamic studies confirmed HSP90 inhibition. Further evaluation of BIIB021 in GIST is warranted.

rsc: a novel oncogene with structural and functional homology with the gene family of exchange factors for Ral.

A novel oncogene, rsc (rabbit squamous cell carcinoma), has been identified from a DMBA-induced rabbit squamous cell carcinoma using gene transfer and the nude mouse tumorigenesis assay. A full-length cDNA has been isolated and sequenced. rsc has potent tumorigenic activity in nude mice (latency <4 weeks), but does not induce focus formation or anchorage independent growth. The oncogene resulted from the fusion of rHR 23A (a rabbit homologue of yeast Rad 23) with a member of the ral-GDS family which we named rgr (ral-GDS related). Deletion analysis demonstrated that the oncogenic potential resides in the Rgr portion of the gene. Rgr is 40% identical overall to Ral-GDS, with identity increasing to 72% over a 100 amino acid region of the catalytic domain. Biochemical experiments indicate that Rgr has GTP/GDP exchange activity for Ral, providing evidence that this pathway is associated with tumorigenesis. The linkage between the Ral pathway and tumorigenesis by a molecule in the Ral-GDS gene family (Ral-GDS being a known effector for Ras) will open the way for the characterization of this pathway and provide an important tool to understand its biological function.