Purpose: Venous and lymphatic malformations (VMs and LMs) carry postzygotic somatic mutations in malformation endothelial cells (VMECs, LMECs), leading to PI3K/AKT/mTOR hyperactivation.Off-label treatment with sirolimus, an mTOR inhibitor, and alpelisib, a PI3K inhibitor, has been effective, confirming the validity of a pharmacological approach in the treatment of VMs and LMs. However, treatment response is incomplete. We hypothesized that a high-throughput screen (HTPS) against VMECs and LMECs carrying PIK3CA mutations would identify novel candidate drugs to be used in the treatment of VMs and LMs. Methods: FDA-approved/in-clinical-trial drugs at 1uM concentration, and percent viability relative to vehicle-treated controls was determined at 48 hours. A dose response viability assay was performed to compare effective inhibitors to sirolimus and alpelisib. In a xenograft mouse model, VMECs (n=2) and LMECs (n=2) carrying PIK3CA variants were resuspended in Matrigel, which were subcutaneously implanted into the flanks of nude mice. Malformations were allowed to develop for either one week (VMs) or two weeks (LMs) before beginning randomized treatment with either oprozomib (a proteasome inhibitor) or a vehicle. Oprozomib was administered by mouth at 30mg/kg 2x per week for 4 weeks. Implants were harvested and analyzed histologically and by immunofluorescence, and relative vascular areas to total implant areas were measured. Statistical analysis was performed with one-way ANOVA with posthoc Tukey’s HSD as well as student’s t-test. Results: HTPS identified three proteasome inhibitors (PIs) that significantly suppressed VMEC and LMEC viability when compared to sirolimus and alpelisib: carfilzomib, delanzomib, and ixazomib (p<0.0001, ANOVA). In a dose response viability assay, all 6 PIs significantly inhibited VMEC and LMEC at clinically relevant doses (p<0.005, student’s t-test. Oprozomib was well-tolerated by mice in the xenograft model and effectively inhibited VM and LM development. When oprozomib-treated implants were compared to controls, LM channel sizes were significantly decreased (p<0.02), and relative vascular area of VM implants were also significantly decreased (p<0.007, student’s t-test). Conclusion: We identified a novel class of drugs, proteasome inhibitors, that significantly inhibited VMEC and LMEC viability in vitro and in vivo. These PIs had superior efficacy in VMEC and LMEC inhibition relative to sirolimus and alpelisib. These data demonstrate that proteasome inhibitors may be efficacious in VM and LM treatment. Additional studies are needed to further understand mechanisms of PI efficacy.
Read full abstract