A structure-based virtual screening-driven discovery pharmaceutics platform for cancer therapeutics development: FDA 505(b)(2) repurposing of netupitant as a BMX-targeted nanomedicine using quality-by-design lipid nanoparticles and model-informed PBPK translation

Bone marrow tyrosine kinase on chromosome X (BMX) is an emerging therapeutic target implicated in prostate cancer progression, metastatic disease, and resistance to current therapies; however, no selective BMX-targeted therapy has advanced to clinical development. Building upon our previously established AI-driven BMX discovery platform, this study presents an integrated Discovery Pharmaceutics strategy combining artificial intelligence-assisted drug repurposing, Quality-by-Design (QbD)-guided lipid nanoparticle (LNP) formulation, experimental validation, and physiologically based pharmacokinetic (PBPK) modeling to accelerate clinical translation through the FDA 505(b)(2) regulatory pathway. An structure-based virtual screening workflow comprising 2,115 FDA-approved small molecules identified netupitant as one of the highest-ranked candidates against the BMX catalytic domain, exhibiting a predicted binding affinity of -10.9 kcal/mol together with favorable selectivity over closely related Tec-family kinases. Netupitant demonstrated concentration-dependent inhibition of prostate cancer cell growth with IC 50 values of 15.97 μM in DU145 cells and 7.1 μM in PC3 cells. A Box-Behnken Design optimized Netupitant-loaded lipid nanoparticles (Netupitant-LNPs) with a particle size of 133.9 ± 3.2 nm, polydispersity index of 0.213 ± 0.005, zeta potential of + 30 ± 1.3 mV, and encapsulation efficiency of 97.3 ± 0.7%. Cryogenic transmission electron microscopy confirmed uniformly dispersed spherical nanoparticles with an average diameter of 130 ± 6 nm, consistent with dynamic light scattering measurements. The nanoformulation substantially enhanced therapeutic efficacy, reducing IC 50 values to 1.1 μM in DU145 cells and 0.5 μM in PC3 cells, representing approximately 15-fold and 14-fold improvements, respectively, over free netupitant. To facilitate clinical translation, experimentally determined formulation characteristics were incorporated into a mechanistic Simcyp® Version 25.2 middle-out physiologically based pharmacokinetic model. The exploratory PBPK model compared alternative intravenous dosing scenarios and identified 100 mg once weekly as a candidate regimen for further evaluation; these simulations do not establish a clinical dosing recommendation. Collectively, this work demonstrates that integrating AI-enabled drug repurposing, Discovery Pharmaceutics, nanomedicine engineering, and model-informed drug development provides an effective strategy for accelerating oncology drug development. Netupitant-LNPs emerge as a repurposed nanomedicine candidate with predicted BMX binding while establishing a broadly applicable translational framework for FDA-approved drug repurposing through the 505(b)(2) pathway.

Abstract from PubMed (PMID 42946268). This entry was added automatically by our daily literature monitor because it matches the topics we follow; read the full paper at the original source.