Targeting RET fusions in sarcomas with vepafestinib, a 3rd generation RET inhibitor with superior blood-brain barrier penetration

Background RET fusions are known drivers across multiple cancer types. While the FDA-approved small-molecule RET inhibitors selpercatinib and pralsetinib yield overall response rates of ~50-80%, on- and off-target resistance mutations and CNS disease progression underscore the need for novel therapies. Vepafestinib is a RET inhibitor with enhanced specificity and potency against RET solvent-front (G810) and gatekeeper (V804) mutations. Preclinical models of brain metastases demonstrate superior CNS penetration and anti-tumor activity compared to selpercatinib. Here, we investigated the therapeutic potential of vepafestinib in RET fusion-driven sarcomas. Methods Patient-derived xenograft (PDX) and cell line models of a SPECC1L::RET-driven sarcoma were established from a brain metastasis. Isogenic cell line models expressing SPECC1L::RET were generated via CRISPR-Cas9 genomic editing of human mesenchymal stem cells (HMSC). Xenograft models were developed by implanting either PDX tissue or HMSC-RET cells into the subcutaneous flank of NOD scid gamma (NSG) mice. To model brain metastasis, HMSC-RET cells stably expressing a bioluminescent reporter were implanted into the cerebellum of NSG mice. Protein expression and phosphorylation were assessed by Western blotting. Cell growth was measured using a viability dye. Results We established a PDX (Sarc-01pdx), a patient-derived cell line (Sarc-01cl), and an HMSC line harboring the SPECCL1::RET fusion (HMSC-RET). Expression of SPECCL1::RET was confirmed by RT-PCR and Western blot. Vepafestinib effectively inhibited the growth of Sarc-01cl (IC50=0.09 µM) and HMSC-RET (IC50=0.21 µM), compared to the isogenic control (HMSC IC50=12.7µM), with dose- and time-dependent reductions in the phosphorylation of RET (Y905 and Y1062) and downstream effectors, including AKT and ERK1/2. While selpercatinib and pralsetinib also inhibited the growth of the RET fusion-driven sarcoma cells, they showed more off-target effects in the control HMSC cells relative to vepafestinib.In both PDX and cell line xenograft models, vepafestinib (50 mg/kg BID) caused tumor growth inhibition comparable to selpercatinib (10 mg/kg BID) and pralsetinib (15 mg/kg BID). Upon cessation of treatment, tumor recurrence was not observed in mice treated with vepafestinib, whereas regrowth was detected in 1/5 and 3/5 mice in the selpercatinib and pralsetinb groups, respectively. Intracranially, vepafestinib was significantly more effective than selpercatinib at blocking tumor growth (p=0.042) and improving survival (median: 56 vs 32 days, p=0.0025). Conclusions Our preclinical results support vepafestinib as a promising, CNS-active therapy for RET fusion-driven sarcomas. Vepafestinib is currently being evaluated in the ongoing phase 1/2 margaRET trial for advanced RET-altered solid tumors (NCT04683250).