Optimizing the Cardiotoxicity Profile of Osimertinib via Scaffold Hopping: An Integrated Cheminformatics and Molecular Modelling Approach

Osimertinib, a third-generation covalent EGFR inhibitor widely used to treat EGFR T790M-positive Non-Small Cell Lung cancer (NSCLC), is limited by dose-dependent cardiotoxicity driven by hERG channel inhibition. To design safer analogues, we applied an integrated computational workflow combining ADMET filtering, covalent docking, molecular dynamics, and quantum-chemical analysis. Using ADMETopt and admetSAR 2.0, a scaffold-hopping campaign generated 390 analogues, of these 161 compounds exhibited substantially reduced predicted hERG inhibition (<0.70) and were prioritized for structure-based studies. Covalent docking against double mutant L858R/T790M EGFR TK (PDB ID: 5HCX) using Schrödinger CovDock identified 101 analogues with covalent docking scores below -5.0 kcal·mol -1 , comparable to or better than Osimertinib (-5.79 kcal·mol -1 ) and retaining productive orientation toward Cys797. Top analogue compound 252 demonstrated favourable covalent docking (-9.26 kcal·mol -1 ) and was subjected to a 100-ns molecular dynamics simulation. The complex exhibited lower backbone RMSD, reduced protein fluctuations, and persistent active-site interactions relative to the Osimertinib complex. PCA, DCCM, and FEL analyses confirmed stabilization of EGFR in a single dominant conformational basin, indicative of enhanced dynamic stability. DFT calculations at the B3LYP/6-311G++** level showed compound 252 possessed a favourable HOMO-LUMO gap (0.1416 eV), lower LUMO energy, and higher electrophilicity (ω = 0.115 eV), supporting efficient Michael acceptor reactivity. Overall, compound 252 represents a quantitatively validated cardiotoxicity-optimized EGFR inhibitor, demonstrating improved stability, reactivity, and predicted safety. This study establishes a robust computational framework for designing next-generation covalent kinase inhibitors with reduced off-target cardiotoxicity.