AI is already integrated into our lives in many ways. It’s changing the way we work, search and communicate. And while that kind of AI use may not be for everyone, perhaps one thing we can all agree on is how it is progressing medicine through breakthroughs that may have not happened without it. At least, not as quickly.
Researchers are using AI to understand cancer at a level of detail that would have been difficult to imagine just a decade ago. Rather than treating every patient with the same type of cancer in the same way, scientists are working toward treatments based on the unique genetic makeup of each person’s tumor.
And there are encouraging signs that more personalized cancer treatments could make a real difference. In October, researchers reported promising Phase 3 clinical trial results for an experimental personalized cancer vaccine developed by Moderna and Merck.
When combined with the immunotherapy drug Keytruda, the vaccine reduced the relative risk of melanoma recurrence or death by 42% compared with Keytruda alone.
The vaccine isn’t FDA-approved yet, and the trial doesn’t establish that generative AI invented the treatment. But the results offer an important example of what personalized cancer medicine can achieve.
Meanwhile, AI is helping researchers tackle another part of the challenge: understanding exactly what makes one person’s cancer different from another’s.
AI could help doctors find the right treatment for each patient
One of the biggest challenges in treating cancer is that two people with the same diagnosis can respond very differently to the same medication.
That’s partly because cancer isn’t a single disease. Even tumors that appear similar can have different genetic mutations, making one treatment effective for one patient and less useful for another.
A study within the last six months highlights that instead of relying solely on patterns found across thousands of patients, researchers can increasingly investigate the biological characteristics of a specific person’s tumor.
AI is particularly useful because it can analyze enormous quantities of genetic information and identify relationships that might otherwise be missed.
The National Cancer Institute has already demonstrated one approach. Its PERCEPTION AI tool analyzes gene activity in individual cancer cells to predict whether certain drugs might work.
This is a significant breakthrough since tumors can contain different populations of cells. A drug might successfully attack some while leaving others behind, potentially allowing the cancer to return or develop resistance.
Tools like PERCEPTION could eventually help doctors understand those differences before deciding which treatment to recommend.
Google DeepMind is helping speed up drug discovery
Another significant development involves AlphaFold, the AI system developed by Google DeepMind.
Proteins play a central role in how cells function, including how cancer develops and spreads. Understanding their three-dimensional structures can help scientists determine how potential drugs might interact with them.
Traditionally, identifying thosestructures could take years of laboratory work. AlphaFold can predict many protein structures in minutes, giving researchers a much faster starting point.
Google DeepMind has made predictions for more than 200 million protein structures available to researchers worldwide. And while that doesn’t mean AlphaFold can automatically create a safe or effective cancer drug, scientists still need to determine whether predicted interactions work in laboratory experiments and, eventually, in patients. It can however, possibly help narrow narrow down the possibilities, potentially saving researchers significant time during the earliest stages of drug development.
AI and clinical trials
For patients and families, the questions to ask are whether the advances in AI and personalized medicine will actually lead to better treatments. The recent melanoma vaccine results offer one encouraging example of the broader shift toward individualized care.
The experimental vaccine, calledintismeran autogene, is designed using genetic mutations found in a patient’s tumor. Those mutations help identify cancer-specific proteins that the immune system can learn to recognize.
In the Phase 3 trial involving 1,137 patients whose melanoma had been surgically removed, 82% of those receiving the personalized vaccine alongside Keytruda remained free of recurrence at two years, compared with 70% receiving Keytruda alone
If AI can make analyzing that information faster, more accurate and more accessible, it could help expand the possibilities for personalized medicine.
The direction of the research is encouraging
There is still a considerable gap between discovering a promising treatment and making it available to patients. Even when AI identifies a potential drug or predicts how a tumor might respond, those findings need to be validated. Clinical trials remain essential for demonstrating that a treatment is both safe and effective.
There are also practical challenges, including the costs of genetic testing, access to specialized facilities and ensuring that AI systems perform reliably across different patient populations.
Yet, it’s clear that AI is giving scientists new ways to study cancer, identify possible drug targets and predict which treatments may work best for particular patients. At the same time, personalized therapies are beginning to show meaningful results in large clinical trials.
The goal isn’t to replace oncologists, but to give doctors better information and more treatment options.
- Man sues OpenAI alleging ChatGPT helped his wife try to murder him: it’s the 10th homicide the AI chatbot has been accused of abetting in less than two years
- A few tech leaders are rebranding AI as ‘superintelligence’ — calling it that doesn’t make it accurate
- A new bill could ban powerful AI in the US — here’s what it means for ChatGPT
