← Back to Home

Iambic submits IND for AI-designed cancer drug candidate IAM217, exploring KIF18A and brain metastasis treatment

A small-molecule drug targeting a vulnerability in cancer cell division is seeking to enter human trials. IAM217 has shown signals of antitumor activity in the brain in preclinical models, but its ability to cross the human blood–brain barrier and its therapeutic benefits still require clinical validation.

By SURL BioNews

For a cancer drug to treat lesions in the brain, it must not only inhibit tumors but also reach sufficient concentrations at its site of action. Iambic Therapeutics has included brain penetration among the design goals for IAM217, which it is advancing. This KIF18A inhibitor, developed with the help of artificial intelligence, seeks to link a vulnerability in cancer cell division with the need for brain metastasis treatment.

Iambic announced on October 5 that it had submitted an investigational new drug application (IND) to the U.S. Food and Drug Administration (FDA), with plans to begin a Phase 1/2 trial evaluating IAM217 in triple-negative breast cancer, ovarian cancer, and other solid tumors with chromosomal instability. A summary of the Business Wire announcement republished by San Diego Biotechnology Network that same day also confirmed the submission and the main intended indications. Human trials still await FDA authorization to begin; submitting an IND does not mean the drug has received marketing approval.

KIF18A is a motor protein involved in cell division that helps align chromosomes. The company's development rationale is that cancer cells with chromosomal instability rely more heavily on it to complete division, potentially creating a vulnerability that can be targeted. If normal cells tolerate KIF18A inhibition better, this could provide a therapeutic window, but this selectivity still needs to be tested in humans.

AI's role here is molecular design. According to Iambic, in the absence of a high-resolution experimental structure of KIF18A bound to an inhibitor, the team used NeuralPLexer to provide structural clues, then used property prediction tools to balance inhibitory potency, oral absorption, drug interaction risk, and brain penetration. These goals involve different molecular properties, and whether they have been achieved cannot be determined through computer predictions alone.

The main efficacy signal disclosed by the company comes from preclinical brain metastasis models: IAM217 shrank established intracranial tumors by approximately 90%. This result supports further exploration, but the announcement did not fully present the model setup, sample size, control results, or duration of response. It therefore cannot be used to predict patients' treatment responses or confirm that the drug can reach effective concentrations in the human brain.

The available information also has limitations in terms of its sources. The content published by San Diego Biotechnology Network is a press release summary, rather than a separate set of independent experiments; the brain penetration and antitumor findings remain primarily company-disclosed data. Although there is a specific description of how AI influenced candidate molecule selection, directly comparable evidence of its advantages over other design approaches is still lacking.

If the IND receives authorization to proceed, IAM217 will become Iambic's second development program to enter clinical trials. The next priorities are to identify doses that humans can tolerate, confirm drug exposure and preliminary antitumor activity, and then determine which patients might benefit. Expectations for brain metastasis treatment will also move from tumor shrinkage in models to more rigorous clinical testing.

References

  1. Iambic Therapeutics
  2. San Diego Biotechnology Network