AI-Designed Drug Candidates Gain Recognition at ASCO 2024: Highlights from Four Clinical-Stage Molecules and an AI-Engineered Cancer Vaccine

May 31 to June 4, 2024American Society of Clinical Oncology (ASCO)The annual meeting was grandly held in Chicago.

As the world's largest, highest academic level, and most authoritative clinical oncology conference, this year'sASCOBrings together many world-class oncology experts.

In recent years, the medical community's attitude towards embracing new AI technologies has become increasingly clear.The conference also featured the presence of several AI pharmaceutical companies.

WeMore than 5,000 abstractsSelected and organizedAmong whichDrug molecules designed by AI,AI-Designed Antigen VaccineAndArtificial Intelligence and Machine Learning Models。

What are the latest advances AI pharmaceutical companies have made against cancer, the number one killer of human health? Which AI drug molecules have seen new progress?

Four AIDesigned Drug Molecules

Chimeric Degrader AC0699

R&D Company:Accutar Biotech

Clinical Stage:Phase 1

This is a potential best-in-class drug, belonging to a new class of chimeric degraders with a novel mechanism, targeting breast cancer by bringing ERα into close proximity with the E3 ligase for binding, ubiquitination, and degradation of ERα.

Compared with traditional Selective ER Degraders (SERD), it has higher specificity and more complete target blockade. The preliminary data from the Phase 1 trial are positive, and the Phase 2 study will begin recruitment in early 2024.

Founded in 2015Iceland Spar BioIt is an old-brand AI pharmaceutical company, with independent research and development of several modules such as the end-to-end retrosynthesis system ChemiRise, the docking platform Orbital, and the ADME property prediction platform Chemi-Net.

AC0699 demonstrated breakthrough safety and efficacy in Phase 1 trials, preliminarily provingBest-in-classPotential, looking forward to the clinical performance in patients with ESR1 mutations.

Humanized Igg1 Monoclonal Antibody AU-007

R&D Company:Aulos Bioscience

Clinical Stage:Phase 1/2

This isThe World's First AI-Designed Humanized Monoclonal Antibody to Enter Clinical Trials, aiming to utilize interleukin-2 (IL-2) to eradicate solid tumors in patients with unresectable locally advanced or metastatic cancer.

Unlike existing therapies, the drug expands the therapeutic window by redirecting IL-2 to CD8+ T and natural killer cells, which can prevent the negative feedback loop caused by other IL-2-based treatments and shift the immune system towards activation rather than suppression.

Trial data show that the drug has durable anti-tumor activity across multiple solid tumor types. Updated clinical data are expected to be available in the second half of this year, and a Phase 2 trial is planned.

HPK1 Inhibitor NDI-101150

R&D Enterprises:Nimbus Therapeutics

Clinical Stage:Phase 1/2

This is a novel oral small-molecule Hematopoietic Progenitor Kinase 1 (HPK1) inhibitor. In the phase 1/2 trial, 16.7% of patients benefited from monotherapy, with one patient with renal cell carcinoma (RCC) achieving complete remission.

Although a 16.7% clinical benefit rate may seem modest, Nimbus is highly optimistic and stated that these data further demonstrate HPK1 as a "differentiated next-generation immunotherapy target" for solid tumors.

Nimbus TherapeuticsFounded in 2009 and co-founded by computational chemistry pioneer Schrödinger, this drug is the company's only clinical candidate.

HER2 Inhibitor IAM1363

R&D Enterprises:Iambic Therapeutics

Clinical Stage:Phase 1/1b

This is a selective and brain-penetrant small molecule inhibitor targeting the HER2 protein, discovered through Imbic's AI-driven platform. It aims to expand the therapeutic index compared to existing HER2 inhibitors and avoid toxicity caused by off-target EGFR inhibition.

In preclinical studies, the drug's selectivity for HER2 isMore than 1000 times of EGFR, with excellent pharmacokinetics and safety, preferential tumor enrichment, and high exposure in the central nervous system.

In the recently released Phase I/IbTestIn China, the drug demonstrated efficacy in HER2 tumor models (including intracranial tumor models) compared to TKIs and HER2 ADCs.Showed good efficacy and tolerability.

An AIDesigned Antigen Vaccine

Star AI Pharmaceutical EnterprisesEvaxionPhase 2 Study Data of EVX-01, a Peptide-Based Cancer Vaccine, Reported at This Conference.

ERV (Endogenous Retrovirus) is a component of the "dark matter" of the human genome, accounting for about 8% of the human genome.

Recent studies show that ERVs are usually overexpressed in cancer but not in healthy populations, making them visible to the immune system and thus a promising target for cancer vaccines.

The vaccine, designed by an in-house AI platform, demonstrated safety and good tolerability when used in combination with Merck's PD-1 pembrolizumab for melanoma treatment.

Responses of CD8+ and CD4+ T cells to corresponding vaccine peptides were detected in the data, and it was found that both the number of immunogenic peptides and the PIONEER epitope quality score could predict favorable CR and longer PFS.

Objective responses were observed in patients with metastatic melanoma at all three dose levels, and all patients at the highest dose level achieved CR.

EVX-01-Specific PBMC-Derived T Cell Responses Analyzed by Elispot and ICS

The achievement is supported byAI-Immunology™, which integrates the company's original threeAI-Driven Immunology PlatformThe integration of PIONEER™, EDEN™, and RAVEN™, after being trained through simulated immune systems, can economically and efficiently design and develop unique immunotherapies.

The data from the current Phase 2 study confirms the findings from Phase 1, reaffirming the AI platform's ability to precisely select treatment-related vaccine targets.

The Preferred Efficiency-Boosting Solution for MNCs and Large Hospitals

In addition to self-developed drugs and therapies, AI CROs were also highly active at this conference.Artificial intelligence has gained increasing recognition from multinational pharmaceutical companies and medical institutions.

For exampleOnc.AiAndPfizerCollaboration to develop and validate a deep learning imaging tool using real-world datasets (RWD).

This tool can inform clinical decision-making, such as helping guide whether concurrent chemotherapy may not be necessary, achieving more than three times the efficiency.

In addition,Predictive OncologyAndMagee Women's Hospital of the University of Pittsburgh, USAThe results of the collaboration were also disclosed at the meeting.

This AI drug discovery company has developed a multi-omics machine learning model using artificial intelligence technology, achieving excellent predictions for both short-term and long-term survival of ovarian cancer patients. The best-performing multi-omics model demonstrates superior predictive power compared to any isolated feature set.

AI New Models: A Bursting Boom

—The End—

Recommended Reading