AI-Driven Study Identifies (Z)-Endoxifen as Promising Therapeutic Candidate for Glioblastoma
Key Insights
Insilico Medicine (search) and Atossa Therapeutics published a comprehensive AI-enabled study in Nature Scientific Reports identifying (Z)-endoxifen as a potential treatment for glioblastoma multiforme (search), one of the deadliest brain tumors.
The research utilized Insilico's PandaOmics platform to evaluate over 900 cancer indications, revealing that endoxifen could counteract key pathways driving tumor growth and treatment resistance in glioblastoma (search).
Laboratory validation demonstrated that (Z)-endoxifen significantly suppressed GBM cell proliferation and induced apoptosis, showing greater cytotoxic activity than high-dose temozolomide in vitro.
Insilico Medicine (search) and Atossa Therapeutics have published groundbreaking research in Nature Scientific Reports demonstrating the potential of (Z)-endoxifen as a therapeutic candidate for glioblastoma multiforme (search) (GBM), one of the most aggressive and deadly brain tumors in adults. The peer-reviewed study represents one of the most comprehensive AI-enabled analyses to date exploring new therapeutic opportunities for a disease with a devastating five-year survival rate of roughly 4%.
AI-Powered Drug Repurposing Breakthrough
The research utilized Insilico's proprietary PandaOmics platform to systematically evaluate more than 900 cancer indications, seeking new oncology applications for (Z)-endoxifen, an active metabolite of tamoxifen already clinically active in hormone-resistant breast cancer (search). Through this comprehensive analysis, GBM emerged as a top candidate for further investigation.
"This collaboration with Insilico Medicine (search) provides a whole new indication in which we might explore the utility of endoxifen, and potentially significant new opportunities to address an extremely underserved set of cancer patients," said Steven Quay, M.D., Ph.D., CEO of Atossa Therapeutics.
The AI-driven analysis identified more than 1,400 genes shared between GBM tumors and endoxifen-treated cells, revealing strong reversal of biological programs that drive tumor growth and treatment resistance. The platform combined differential expression, pathway enrichment, protein-protein interaction networks, mechanistic natural language processing, and disease unmet-need modeling to rank GBM among the highest-opportunity indications.
Mechanistic Insights and Pathway Analysis
Single-cell sequencing analysis pinpointed key malignant-cell genes linked to poor survival and aggressive GBM subtypes, all of which were downregulated by endoxifen treatment. The study revealed that endoxifen was predicted to counteract pathways associated with uncontrolled proliferation, inflammation, metabolic dysregulation, and aggressive tumor behavior.
(Z)-Endoxifen acts through both estrogen-dependent and estrogen-independent mechanisms and has long been considered promising for broader oncology applications, though it had not been systematically explored in GBM until this study.
Laboratory Validation Confirms Computational Predictions
The experimental validation closely matched computational predictions, providing strong evidence for endoxifen's therapeutic potential. In vitro studies demonstrated that (Z)-endoxifen significantly suppressed GBM cell proliferation and induced apoptosis, showing greater cytotoxic activity than high-dose temozolomide and enhanced effects when used in combination.
In vivo studies further confirmed that endoxifen was well tolerated across all tested doses, supporting its safety profile for potential clinical development in GBM patients.
Implications for Drug Discovery Efficiency
The collaboration showcases the transformative potential of AI-powered drug discovery. Alex Zhavoronkov, Ph.D., Founder and CEO of Insilico Medicine (search), noted that "this publication represents one of the early fruits of our joint research, and much of our collaborative work remains unpublished. We are hopeful that these efforts will ultimately translate into meaningful therapeutic programs."
Insilico's AI-driven approach has dramatically improved drug discovery efficiency. While traditional early-stage drug discovery typically requires 3 to 6 years, from 2021 to 2024 Insilico nominated 20 preclinical candidates, achieving an average turnaround from project initiation to preclinical candidate nomination of just 12 to 18 months per program, with only 60 to 200 molecules synthesized and tested in each program.
Expanding Therapeutic Horizons
The study represents a significant step toward addressing the urgent unmet need in glioblastoma (search) treatment. Dr. Quay emphasized that the research extends beyond their traditional focus on women's health, noting potential applications in "other highly ranked indications, including but not limited to Duchenne muscular dystrophy (search) (DMD) and multiple gynecologic-related cancers for which (Z)-Endoxifen may have excellent therapeutic outcomes."
The multi-omic analyses and experimental findings highlight endoxifen as a promising therapeutic candidate for GBM and demonstrate how AI-powered discovery can reveal new opportunities for drug repurposing and cancer treatment innovation, potentially offering hope to patients facing one of oncology's most challenging diagnoses.