Evogene, Queensland University of Technology Form Collaboration on AI Cancer Therapeutics

Evogene Ltd. (Nasdaq: EVGN; TASE: EVGN), a Rehovot, Israel-based computational chemistry company, and the Queensland University of Technology (QUT) in Australia announced a research collaboration to apply generative AI cancer therapeutics approaches to the discovery of novel small-molecule drug candidates for chemotherapy-resistant non-small cell lung cancer (NSCLC). The deal, disclosed on February 17, 2026, pairs Evogene's proprietary ChemPass AI platform with biological findings from the lab of Dr. Mark Adams, a cancer genomics researcher in QUT's School of Biomedical Sciences and Faculty of Health. Financial terms were not disclosed.

The Evogene QUT collaboration centers on a newly identified enzymatic detoxification pathway that Dr. Adams' group has linked to cisplatin resistance in NSCLC cells. The specific molecular target was not named in the announcement. Evogene will deploy its ChemPass AI engine to generate and iteratively optimize novel small-molecule inhibitors of this pathway, while QUT will contribute biological validation and experimental data to refine the computational models. No drug candidates have been identified or named; the program is at the discovery stage.

Chemotherapy Resistance in Lung Cancer Remains a Persistent Clinical Problem

The collaboration addresses a well-documented gap in NSCLC treatment. Cisplatin, a platinum-based chemotherapy agent, remains a cornerstone of NSCLC management, yet intrinsic resistance is observed in 60-70% of treated patients, according to figures cited in the press release referencing Hendriks et al. in Nature Reviews Disease Primers (2024). An additional 30-40% of patients receiving targeted therapies fail to respond upfront, per Passaro et al. in ESMO Open (2020). Comparable rates of intrinsic and acquired resistance are reported for immunotherapy, whether administered as monotherapy or in combination with chemotherapy, per Schoenfeld et al. in Annals of Oncology (2021).

No drugs are currently approved with a specific indication for overcoming cisplatin resistance. When platinum-based regimens fail, clinical practice typically involves switching to alternative systemic therapy classes rather than employing mechanism-directed re-sensitization strategies. The absence of agents that selectively target the biology of platinum resistance has been noted in the literature, including work by Exploiting ROS and metabolic differences to kill cisplatin-resistant lung cancer (Hagen et al., 2017).

The biological rationale for the collaboration rests on preclinical evidence that cisplatin-resistant NSCLC cells upregulate cellular detoxification capacity, particularly through NADPH-dependent redox pathways. Published studies have demonstrated that inhibiting enzymes in the oxidative pentose phosphate pathway, such as glucose-6-phosphate dehydrogenase (G6PD) and 6-phosphogluconate dehydrogenase (6PGD), can restore cisplatin sensitivity in resistant lung cancer cell lines. The exact relationship between these published targets and the "novel druggable cellular detoxification pathway" referenced in the announcement was not specified.

AI-Driven Drug Discovery Applied to a Novel Resistance Target

The operational structure of the collaboration follows a closed-loop design model. Evogene's ChemPass AI platform will generate candidate molecules computationally, prioritizing compounds with inhibitory potential against the target pathway and favorable drug-like properties including metabolic stability, solubility, and permeability. Dr. Adams' laboratory will then test these computationally designed molecules in biological assays, and the resulting data will be fed back into the generative model for further rounds of multi-parameter optimization.

This iterative approach is particularly relevant here because the target pathway, as described, is newly characterized and lacks existing chemical matter or known pharmacophores. Traditional high-throughput screening campaigns rely on existing compound libraries that may not contain suitable starting points for a novel target. Generative AI-driven drug discovery platforms can propose de novo chemical scaffolds without dependence on prior structure-activity relationship data, a capability that Evogene positions as a core differentiator of its computational chemistry cancer approach.

The collaboration did not disclose any timeline for candidate selection, preclinical development milestones, or plans for regulatory filings. No IND-enabling studies or clinical trial designs were referenced. Geographic rights, commercialization terms, royalty structures, and milestone payments were not mentioned, consistent with the early-stage, academic nature of the partnership.

Competitive Landscape for Detoxification-Targeting Agents

Several programs have attempted to exploit redox and detoxification biology to overcome platinum resistance, with mixed clinical results.

RRx-001 (nibrozetone), developed by EpicentRx, is the most mechanistically proximate competitor. The compound depletes glutathione, inhibits G6PD activity, and generates reactive oxygen species. It is in Phase III evaluation in the REPLATINUM trial for platinum-resistant small cell lung cancer, with earlier Phase II data showing platinum re-sensitization in heavily pretreated patients. RRx-001 represents the most advanced clinical validation of the G6PD/glutathione depletion approach to chemo-sensitization.

NOV-002, a formulation of oxidized glutathione developed by Novelos Therapeutics, reached Phase III in first-line advanced NSCLC in combination with cisplatin and paclitaxel but failed to meet its primary endpoint. The program has been dormant since approximately 2010. The failure highlighted the difficulty of translating redox modulation into clinical benefit in a broad, non-biomarker-selected NSCLC population.

Telaglenastat (CB-839), a selective glutaminase inhibitor developed by Calithera Biosciences in collaboration with Merck, was tested in the KEAPSAKE Phase II trial in KEAP1/NRF2-mutant NSCLC combined with pembrolizumab and chemotherapy. The trial did not meet its primary endpoint, suggesting that indirect reduction of glutathione precursor supply may be insufficient and that more proximal targets within the detoxification pathway could offer greater potency.

Auranofin, a gold-containing thioredoxin reductase inhibitor originally approved for rheumatoid arthritis, is in Phase I/II evaluation in multiple cancers including lung, with preclinical evidence of cisplatin re-sensitization via disruption of NADPH-dependent redox defense. Buthionine sulfoximine (BSO), an irreversible inhibitor of the rate-limiting enzyme in glutathione synthesis, has been explored in Phase I/II settings but has been limited by therapeutic window concerns.

No selective, clinical-stage inhibitor of G6PD or 6PGD currently exists. If the Evogene-QUT program targets one of these enzymes specifically, it would occupy a differentiated position in a space where the biology has been repeatedly validated but clinical translation has proven difficult.

Evogene Cancer Therapeutics Ambitions and Deal Context

The collaboration represents Evogene's entry into oncology drug discovery. The company's ChemPass AI platform was originally developed for applications across both pharmaceutical and agricultural chemistry. Evogene is a publicly traded small-cap company; as of the announcement date, it trades on both Nasdaq and the Tel Aviv Stock Exchange.

For QUT, the partnership provides access to computational chemistry cancer capabilities that could accelerate translation of Dr. Adams' laboratory findings into optimized chemical leads. Dr. Adams characterized the collaboration as an opportunity to leverage AI-driven technology to move "from research to real-world outcomes."

The deal structure — a research collaboration between a publicly traded technology company and a university research group, with no disclosed financial terms — is typical of early-stage academic-industry partnerships. Any future licensing, commercialization, or development agreements would presumably be negotiated separately, contingent on the scientific output of the initial collaboration. No options, territorial rights, or equity arrangements were disclosed.

The collaboration will need to demonstrate that computationally designed inhibitors of the target pathway can achieve sufficient potency and selectivity in cellular assays, followed by in vivo proof-of-concept, before the program could advance toward IND-enabling studies. Given the discovery-stage nature of the work and the absence of named candidates, clinical evaluation remains distant.

---

Spot something wrong? Report an issue with this article