Ceralasertib

kt-5000AI delivers potent and selective ATR inhibition, CNS-oriented properties and preclinical tumour-growth control, advancing a capital-efficient drug discovery program toward candidate selection VANCOUVER, British Columbia, October 22, 2026 (GLOBE NEWSWIRE) – Rakovina Therapeutics Inc. (TSX-V: RKV) (FSE: 7JO0) (OTCMKTS: RKVTF), a biopharmaceutical company advancing innovative cancer therapies through AI-powered drug discovery, today announced significant preclinical progress from its kt-5000AI program, a collaboration with Vancouver-based Variational AI Inc. (“Variational AI”) focused on the discovery and development of novel, brain-penetrant inhibitors of ataxia telangiectasia and Rad3-related protein kinase (ATR). Rakovina has advanced the program from target specification through lead generation and five cycles of iterative lead optimization in approximately two years, using Variational AI’s Enki™ generative AI platform together with focused experimental validation. The program has progressed from target specification to an advanced lead-optimization stage in approximately two years, with 122 compounds synthesized program-wide and direct external discovery expenditure of approximately US$1 million. By comparison, a conventional program to the same stage would typically synthesize 600–1,700 compounds over roughly 4.5 years at a modelled direct external cost of US$7.6 million, demonstrating the potential for AI-enabled drug discovery to compress timelines and concentrate capital on experiments designed to generate decision-enabling data. “ATR is one of the best-validated targets in the DNA damage response. Eight ATR inhibitors have entered the clinic, and none of them reaches the brain at levels that could support treating CNS tumours or brain metastases,” said Dr. Mads Daugaard, President and Chief Scientific Officer of Rakovina Therapeutics. “That is not an accident of chemistry; the hinge-binding features that make these molecules potent are the same features that keep them out of the brain. Patients with glioblastoma, medulloblastoma and brain metastases from breast, lung and melanoma have no access to this class of medicine, and radiation, the standard of care for these tumours, is exactly the setting where ATR inhibition should add the most. Brain penetration was therefore the defining requirement of the kt-5000AI program from day one. We now have multiple compounds demonstrating the potency, selectivity, metabolic stability and CNS-oriented properties required to pursue a differentiated, brain-penetrant ATR program. Our work with Variational AI has demonstrated that generative AI can help us explore novel chemical space while optimizing the properties that matter most.” “Enki™ compressed hit-finding/ID and hit-to-lead into a single generative AI design cycle: 35 compounds made, four IP-clear lead series, and a brain-penetrant compound in the first round,” said Handol Kim, Co-Founder and CEO of Variational AI. “From there, with each design-make-test-analyze (DMTA) cycle, the program’s own experimental results were fed back into the Enki™ model, so every round of optimization started from better molecules than the last. The lead optimization campaign ran five cycles with only 87 compounds synthesized to reach sub-nanomolar potency, efflux ratios near unity and a tenfold gain in metabolic stability. Typical lead optimization campaigns require 10’s to over 100 DMTA cycles generating thousands of compounds that do not work.” “For a company like Rakovina, AI is how we take cost and time out of drug discovery, and that is a strategic advantage, not an accounting exercise,” said Kim Oishi, Chief Executive Officer of Rakovina Therapeutics. “Instead of synthesizing and testing more than a thousand compounds over four or five years, Enki™ designed molecules that already satisfied the full target product profile, so we made 122 and spent about US$1 million of direct external capital to reach candidate-ready in roughly two years. Our cost per compound is in line with conventional rates; the saving comes from making far fewer molecules, and from getting decision-enabling data in weeks rather than months. That is what allows an emerging biotechnology company to advance a program, and to keep more of the value with our shareholders.” Building a Differentiated ATR Asset Around Brain Penetration ATR is a central regulator of the DNA damage response and an important vulnerability in tumours characterized by elevated replication stress and DNA-repair defects. While several ATR inhibitors have entered clinical development, none has been shown to reach the brain at levels that would support the treatment of CNS tumours. Rakovina and Variational AI established CNS penetration as a core design requirement from the outset of the kt-5000AI campaign. The initial campaign produced multiple novel lead series. In pharmacokinetic studies in mice, a first-generation compound reached levels in brain tissue equivalent to those in plasma. Subsequent optimization has focused on transferring this CNS-penetrant property into increasingly potent and developable chemical series. Multiple Indicators of a Developable Drug Profile The kt-5000AI program has progressed through successive design-make-test-analyze cycles, with resulting compounds demonstrating: Potent ATR inhibition:Multiple optimized compounds reaching single-digit nanomolar and sub-nanomolar potency in a qualified biochemical assay. Strong selectivity:Substantial selectivity over other PIKK-family members such as ATM, DNA-PK, PI3Kα and mTOR. Improved metabolic stability:Intrinsic clearance substantially improved through lead optimization. Favourable permeability and efflux:Apparent permeability consistent with CNS drug-discovery objectives and efflux ratios approaching unity. CNS exposure:Brain penetration established in first-generation compounds within the program’s novel chemical space. Studies to characterize CNS exposure in the optimized series are ongoing. In vivo antitumour activity:An earlier kt-5000AI compound demonstrated tumour-growth control comparable to the clinical-stage ATR inhibitor ceralasertib in a prostate cancer xenograft model, while maintaining favourable tolerability observations. These findings build upon data previously presented by Rakovina and its collaborators at the AACR-NCI-EORTC International Conference on Molecular Targets and Cancer Therapeutics 2025, the Society for Neuro-Oncology 2025 Annual Meeting and the 2026 AACR Annual Meeting. Advancing Toward Development-Candidate Selection The kt-5000AI program remains in preclinical development, with lead optimization continuing through the end of 2026. Planned work includes: Selection of lead and backup compounds Expanded pharmacokinetic characterization Formulation and solubility studies Safety and metabolic profiling Scale-up Other studies required to support IND-enabling development The current results represent preclinical drug-discovery and development progress and do not establish clinical efficacy or safety in humans. kt-5000AI has not entered clinical trials, and a final development candidate has not yet been formally nominated. Rakovina believes the combination of potent and selective ATR inhibition, CNS-oriented drug properties, demonstrated preclinical activity and multiple optimized chemical series provides a strong foundation for continued development and potential pharmaceutical partnering. Joint Webinar Rakovina will host a joint webinar with Variational AI on November 5th at 9:30AM PST to discuss the kt-5000AI collaboration and the program’s progress to date, including the results described in the case study. The webinar will feature Dr. Mads Daugaard, President and Chief Scientific Officer of Rakovina Therapeutics, and Handol Kim, Co-Founder and CEO of Variational AI. Registration details are available at https://us02web.zoom.us/meeting/register/mxBaSUVySR-gz86NaoMSEQ#/registration About the kt-5000AI Program Rakovina’s kt-5000AI program is focused on the discovery and preclinical development of next-generation ATR inhibitors for cancers associated with DNA damage response vulnerabilities, including tumours affecting the CNS. The program is being advanced with Variational AI using the Enki™ generative AI platform. The collaboration combines AI-enabled molecular design, experimental validation and iterative lead optimization to identify novel compounds with optimized potency, selectivity, pharmacokinetic and CNS-penetrant properties. Rakovina retains control over compound selection and development strategy. A detailed case study describing the kt-5000AI program, published by Variational AI, is available at https://variational.ai/news/case-study. About Variational AI Variational AI is redefining the unit economics of drug discovery for better patient outcomes. Founded in 2019 and headquartered in Vancouver, the company has built Enki™, a generative AI platform based on a proprietary foundation model. Enki™ rapidly generates novel, potent, selective, and synthesizable lead-like compounds across 873 pretrained drug targets, including kinases, GPCRs, proteases, and ion channels. With Enki™ 4, the platform extends beyond small molecules to degraders, PROTACs, molecular glues, and ADC payloads. Enki™ supports drug discovery teams with lead generation without relying on high-throughput screening or traditional libraries, eliminating Hit ID and Hit-to-lead. In lead optimization, Enki™’s scaffold-constrained generative engine helps biotech and pharma partners, including Merck/MSD, reach candidate compounds in fewer design-make-test cycles than conventional approaches. More information: https://variational.ai/ About Rakovina Therapeutics Inc. Rakovina Therapeutics is a biopharmaceutical research company focused on the development of innovative cancer treatments. Our work is based on unique technologies for targeting the DNA-damage response powered by Artificial Intelligence (AI) using validated, proprietary platforms. By using AI, we can review and optimize drug candidates at a much greater pace than ever before. The Company has established a pipeline of distinctive DNA-damage response inhibitors with the goal of advancing one or more drug candidates into human clinical trials in collaboration with pharmaceutical partners. Further information may be found at www.rakovinatherapeutics.com. Neither TSX Venture Exchange nor its Regulation Services Provider (as that term is defined in policies of the TSXV) accepts responsibility for the adequacy or accuracy of this release. For Further Information Contact: Rakovina Therapeutics Inc. Kim Oishi CEO Kim.oishi@rakovinatherapeutics.com Mark Mitchell IR@rakovinatherapeutics.com Variational AI Inc. Valentin Beuchillot Director, Business Development & Marketing valentin@variational.ai Notice Regarding Rakovina Therapeutics Forward-Looking Statements: This release includes forward-looking statements regarding the company and its respective business, which may include, but is not limited to, statements with respect to the proposed business plan of the company and other statements. Often, but not always, forward-looking statements can be identified by the use of words such as “plans,” “is expected,” “expects,” “scheduled,” “intends,” “contemplates,” “anticipates,” “believes,” “proposes” or variations (including negative variations) of such words and phrases, or state that certain actions, events, or results “may,” “could,” “would,” “might,” or “will” be taken, occur, or be achieved. Such statements are based on the current expectations of the management of the company. The forward-looking events and circumstances discussed in this release may not occur by certain specified dates or at all and could differ materially as a result of known and unknown risk factors and uncertainties affecting the company, including risks regarding the biopharmaceutical industry, economic factors, regulatory factors, the equity markets generally, and risks associated with growth and competition. Although the company has attempted to identify important factors that could cause actual actions, events, or results to differ materially from those described in forward-looking statements, there may be other factors that cause actions, events, or results to differ from those anticipated, estimated, or intended. No forward-looking statement can be guaranteed. Except as required by applicable securities laws, forward-looking statements speak only as of the date on which they are made, and the company undertakes no obligation to publicly update or revise any forward-looking statement, whether as a result of new information, future events, or otherwise. The reader is referred to the company’s most recent filings on SEDAR+ for a more complete discussion of all applicable risk factors and their potential effects, copies of which may be accessed through the company’s profile page at www.sedarplus.ca.