We have conducted several studies to identify the mechanism of action of rigosertib in Ras-mutated neuroblastoma and rhabdomyosarcoma. Several lines of evidence suggest that rigosertib is not functioning as a Ras mimetic and is instead functioning as a microtubule destabilizing agent in these cell types. First, rigosertib is equally as effective in pediatric cancer cell lines with wild type Ras as it is in pediatric cancer cell lines with mutant Ras. Second, rigosertib induces an M-phase arrest, while inhibitors of Ras signaling generally cause an arrest in G1. We further show that treatment with rigosertib induces mitotic spindle defects and decreases tubulin acetylation, consistent with a decrease in microtubule stability. However, rigosertib is ineffective in xenograft models of rhabdomyosarcoma and neuroblastoma, primarily due to poor tumor penetration of the drug. Rigosertib has different phospho-proteomic effects on pediatric cancer cell lines of different histologies. In particular, rigosertib induces ERK and AKT phosphorylation in rhabdomyosarcoma cells but decreases ERK and AKT phosphorylation in neuroblastoma cell lines. We used a nanostring assay to show that in both Ras-mutated rhabdomyosarcoma and neuroblastoma, expression of 4EBP1 at the RNA and protein level is decreased. Current work is aimed at identifying the mechanism by which this decrease occurs. In addition to this preclinical work, we have written a protocol for a phase I trial of rigosertib in pediatric patients. However, due to the lack of efficacy in animal models we will not pursue this clinical trial. Current work is aimed at identifying drugs synergisitic with rigosertib such that we might improve on the efficacy of this drug in pediatric solid tumors.
