Real-world pharmacovigilance and molecular mechanisms of fruquintinib: SRC and STAT3 as potential off-target mediators of proteinuria

Background: Fruquintinib, a highly selective inhibitor of vascular endothelial growth factor receptors (VEGFR) 1-3, has demonstrated significant survival benefits in metastatic colorectal cancer (mCRC). While pivotal clinical trials have characterized its general toxicity profile, the molecular mechanisms driving specific, high-risk adverse events in the real-world setting remain underexplored. Methods: We adopted a "bedside-to-bench" reverse translational strategy to systematically evaluate the safety profile of Fruquintinib. First, we analyzed spontaneous reports from the FDA Adverse Event Reporting System (FAERS) (Q4 2023-Q1 2025) using disproportionality algorithms (ROR, PRR, BCPNN, and MGPS) to detect significant safety signals. Recognizing proteinuria as a clinically critical organ-specific toxicity, we subsequently prioritized it for mechanistic investigation. A "drug-target-disease" network was constructed, followed by molecular docking and 100-nanosecond (ns) molecular dynamics (MD) simulations to elucidate the underlying molecular etiology. Results: A total of 1,632 Fruquintinib-related reports were identified, with 66.2% of adverse events occurring within the first 30 days of therapy. Significant disproportionality signals were detected for myelosuppression (ROR = 29.31, 95% CI: 23.40-36.71), dysphonia (ROR = 21.96, 95% CI: 18.37-26.26), and proteinuria (ROR = 16.92, 95% CI: 11.81-24.23). Network pharmacology analysis mapped the intersection of drug and disease targets, identifying SRC and STAT3 as core hub kinases. Molecular docking revealed that Fruquintinib exhibits favorable theoretical binding affinities to the active pockets of SRC (-7.8 kcal/mol) and STAT3 (-7.4 kcal/mol), surpassing its affinity for several canonical targets. MD simulations further confirmed the rigorous dynamic stability of these complexes (RMSD ∼0.22 nm) and revealed a distinct induced-fit binding mechanism. Conclusion: This study provides the first integration of pharmacovigilance data with biophysical modeling for Fruquintinib. Beyond corroborating established class related safety profiles, we hypothesize a novel dual hit molecular framework for renal toxicity. This clinical association might involve both classical vascular endothelial growth factor receptor blockade and the potential off target structural interactions of SRC and STAT3. These insights underscore the importance of early renal monitoring and offer structural insights into the drug's safety profile.