Cancer treatment has changed dramatically over the past few decades. Surgery, chemotherapy, targeted drugs and immunotherapy have given many patients more options than ever before.
But there is still a difficult problem: cancers can learn how to resist treatment.
That is especially important for older adults, who are more likely to live with several health conditions at once and may have already gone through multiple cancer treatments by the time a new therapy is considered.
Now, researchers are testing an unusual idea: instead of simply discovering another molecule in nature or modifying an existing drug through years of trial and error, could artificial intelligence help scientists design a molecule for a very specific job?
One experimental treatment, imneskibart, formerly known as AU-007, is putting that idea to the test.
A computer-designed antibody enters human trials
Imneskibart is a monoclonal antibody designed with the help of artificial intelligence by Biolojic Design and being developed clinically by Aulos Bioscience.
It is not an approved cancer treatment. It is currently being evaluated in a Phase 1/2 clinical trial involving people with advanced solid tumors, including melanoma and non-small cell lung cancer.
The scientific idea behind it is surprisingly specific.
The drug is designed to bind to a particular part of interleukin-2, or IL-2, a natural immune-system signaling protein.
IL-2 can stimulate immune cells that attack cancer, including certain T cells and natural killer cells. But IL-2 also interacts with regulatory T cells and blood-vessel cells through receptors containing CD25.
That creates a long-standing challenge.
Researchers want IL-2’s immune-activating effects without reproducing the serious toxicities that limited older high-dose IL-2 treatments.
The strange problem with IL-2
High-dose IL-2 once represented an important breakthrough in treating advanced melanoma and kidney cancer. In a small proportion of patients, it produced remarkably durable responses.
But the treatment could also cause severe toxicity, including vascular leak syndrome, in which blood vessels become excessively permeable and fluid moves into surrounding tissues.
That could lead to serious complications and required intensive medical monitoring.
Imneskibart takes a different approach.
Rather than replacing IL-2, the antibody is designed to attach to the portion of IL-2 involved in binding CD25. This is intended to reduce IL-2 signaling through CD25-containing receptors while allowing it to continue engaging the receptor configuration associated with immune-effector cells.
In simple terms, researchers are trying to redirect the signal rather than simply turn it up.
Why this matters for people with advanced cancer
For someone newly diagnosed with cancer, treatment decisions may involve surgery, radiation, chemotherapy, targeted therapy or immunotherapy depending on the type and stage of disease.
But advanced cancer can become more complicated when a tumor stops responding.
Checkpoint inhibitors such as PD-1 or PD-L1 therapies have transformed treatment for several cancers, yet some tumors eventually progress despite these drugs.
That is where experimental approaches such as imneskibart are being investigated.
The current trial includes patients whose cancers have already progressed after previous treatments. The goal is not simply to produce another immune response, but to determine whether changing the balance between immune activation and immune suppression can translate into meaningful tumor control.
What have researchers seen so far?
The newest publicly reported data came from the 2026 American Society of Clinical Oncology (ASCO) Annual Meeting.
In a Phase 2 melanoma cohort, Aulos reported a 33% objective response rate and 67% disease-control rate among 12 evaluable patients receiving an imneskibart-based triplet regimen after progression on earlier checkpoint-inhibitor combinations. The company also reported reductions in regulatory T cells and increases in the CD8-to-Treg ratio.
Those numbers are certainly worth watching.
But they need context.
The group was small, the treatment remains experimental, and the findings came from an ongoing study. A response rate in a small clinical trial cannot tell us whether the treatment will ultimately improve survival for a broad population of cancer patients.
That distinction becomes particularly important when discussing new cancer therapies with older adults and their families.
Promising is not the same as proven.
Could AI change the way medicines are discovered?
Perhaps the most interesting part of imneskibart is not only the drug itself, but how it was designed.
Biolojic Design describes AU-007 as a computationally designed antibody created to bind a predefined region of IL-2. The company says it was the first computationally designed antibody from its platform to enter human clinical trials.
That points toward a broader change in drug discovery.
Artificial intelligence cannot simply tell researchers which medicine will cure a particular cancer. Human biology is far too complicated for that.
But computational tools can help scientists explore protein structures and potential molecular interactions much more efficiently than traditional trial-and-error approaches.
The eventual test, however, remains the same: does the treatment help patients live longer or live better without unacceptable side effects?
Older adults should watch the evidence—not the hype
For older adults living with cancer, experimental immunotherapy can sound especially exciting when standard treatments have stopped working.
But eligibility for a clinical trial depends on many factors, including cancer type, previous treatments, overall health and the specific requirements of the study.
Age alone does not determine whether someone can benefit from a clinical trial, but other health conditions and medications can matter greatly.
And because imneskibart is still investigational, it should not be viewed as an established alternative to approved cancer treatments.
Researchers are still determining its long-term effectiveness and safety.
The bigger story behind the molecule
Imneskibart may or may not become a successful cancer medicine.
What is already changing, however, is the way scientists think about designing medicines.
Instead of starting only with molecules discovered in nature or modifying existing biological compounds, researchers can increasingly use computational models to design proteins around a desired biological function.
That does not eliminate the uncertainty of cancer.
It does, however, give scientists another way to explore it.
For patients and families, the most important question is ultimately not whether a drug was designed by AI.
It is whether that carefully designed molecule can make a real difference in a human life.
For imneskibart, that answer is still being investigated.
Photo by Pavel Danilyuk: https://www.pexels.com/photo/an-elderly-man-controlling-a-robot-8439071/