(P 192) TARGETING PI3K/MTOR MITIGATES RESISTANCE TO CDK4/6 INHIBITION IN RB-PROFICIENT PEDIATRIC AND AYA OSTEOSARCOMAS

Objective: Despite the use of multimodal therapies, up to 30% of children as well as adolescents and young adults (AYA) with osteosarcoma (OS) have metastatic disease and a dismal prognosis. To improve clinical outcomes, there is a critical need to identify efficacious therapeutic regimens for high-risk pediatric and AYA OS patients. To meet this challenge, the Pediatric Precision Genomics Program at Riley Hospital for Children was established to facilitate development of safe molecularly guided treatments for high-risk pediatric and AYA sarcoma patients. Hyperactivation of cyclin-dependent kinases 4 and 6 (CDK4/6) was identified as an actionable signature in a subset of OS patients. CDK4/6 small molecule inhibitors, however, have not been overly promising as a monotherapy in OS due to several mechanisms including cytostasis induction, inactivated or deleted RB, as well as activation of PI3K/mTOR. Targeting PI3K or mTOR alone may not be sufficient to induce an anti-tumor response due to activation of compensatory pathways or a shift in dependency on PI3K or mTOR. Thus, we hypothesized that inhibition of CDK4/6 in combination with a dual inhibition of PI3K/mTOR will be more efficacious than as single-agents in preclinical models of high-risk pediatric and AYA OS.

Methods: To test our hypothesis, we prioritized HT96 PDX which was established from a primary tumor biopsy from a 9-year old male OS patient that had no prior exposure to therapy and a TT2-77 PDX that was generated from a metastatic resection from an 18-year old male OS patient who was heavily pretreated. Both PDXs exhibit CDK4/6 inhibitor sensitive signatures (RB+, CDKN2Anull, CCND3 amplified). PDX tumor fragments were implanted in flanks of donor gender-matched immunodeficient NOD/SCID/IL2Rγ mice. Once tumor volumes reached ~100mm3, mice were randomized and based on dose-finding studies treated with 5-day cycles of daily treatment with the CDK4/6 inhibitor palbociclib (50 mg/kg) +/- the dual PI3K/mTOR inhibitor voxtalisib (50 mg/kg) for 6-8 weeks. For short term pharmacodynamic studies, ~400-500mm3 TT2-77 PDXs were dosed with palbociclib as described above for 5 consecutive days. Subsequently, tumors were collected for kinome and proteomics analyses. Furthermore, data obtained from these analyses were validated by western blots. In vitro studies were conducted using normal human osteoblasts and mesenchymal stem cells as controls as well as human RB+ OS cell lines [G292 (primary), 143B (metastatic), human metastatic OS xenoline (TT2-77; derived from TT2-77 PDX)], and a murine metastatic OS cell line (K7M2). Pediatric OS cell lines were screened using a matrix format to evaluate degree of synergistic growth inhibition of a range of dose ratios for palbocilib +/- voxtalisib. All cell lines and PDXs have been authenticated and pathogen tested. Statistical significance is achieved when p< 0.05.

Results: Palbociclib + voxtalisib induced a synergistic cell-growth inhibition over a wide range of inhibitor dose ratios in RB+ human and murine OS cell lines but not in RB-deficient cell lines. In the TT2-77 PDX samples treated with Palbociclib for 5 days, decreased activity and expression of CDK1 (kinase involved in G2-M phase progression) was evident by kinome profiling and western blot assays. Additionally, increased activity of AXL, a receptor tyrosine kinase that is linked to downstream PI3K pathway activation as well the autophagy marker PIK3C3 was observed. RB1 and MKI67 phosphopeptides as well as the total CDK1 were reduced by 5-day treatment with palbociclib in the TT2-77 PDX samples, thus, confirming modulation of the cell cycle. Results were validated by western blot. Efficacy studies for TT2-77 PDX and HT96 PDX treated with a 6-8 weeks of palbociclib +/- voxtalisib regimen resulted in a significant growth inhibition following combination therapy versus single-agents and vehicle treatment groups (Fig. 1 p< 0.05). The extended dosing regimens in both PDXs were well-tolerated as indicated by preservation of body weights and histological analysis of organs.

Conclusion: This study is of great significance for it provides strong rationale for earlier therapeutic intervention with a combination therapy targeting CDK4/6 and PI3K/mTOR pathways in a subset of OS patients with CDK4/6 hyperactivation. This therapeutic strategy has the potential to minimize disease burden and enhance quality of life in both the un-pretreated and the pre-treated clinical setting.

Final CTOS Abstract Fig. 1.pdf