Top 3 UT MD Anderson abstracts at AACR 2026
April 15, 2026
At the 2026 American Association for Cancer Research (AACR) Annual Meeting, UT MD Anderson experts will showcase their cutting-edge work that impacts patients across the globe.
I’m so proud of all the progress our researchers have made in discovery science and translational research that will be on display. These advancements will bring forth new therapies for our patients and advance our understanding of the biologic processes that ultimately drive the disease.
For this year’s meeting, our researchers will present nearly 220 abstracts. Here, I’ll highlight three of them that hope to lay the groundwork for future breakthrough discoveries.
A new approach to immune checkpoint inhibition may preserve quality of life for some patients with oral lesions
Approximately 5% of the general population have precancerous lesions in their mouth that can carry up to a 36% risk of progression to oral cancer, explains head and neck surgeon Moran Amit, M.D., Ph.D. While many patients undergo surgery to remove these lesions, the risk of recurrence after surgery can be as high as 40%.
“For these patients, the choice has always been between surgery that gradually erodes their ability to speak and swallow, or watchful waiting that leaves them in a state of uncertainty,” says Amit.
Prior research has found that treatment with the immune checkpoint inhibitor nivolumab can reduce the size and the risk of progression of precancerous oral lesions. But when this treatment is delivered through an intravenous infusion, it can come with toxicities and side effects.
On April 21, at the AACR Annual Meeting, Amit will present findings from a Phase 1 trial that examined injecting nivolumab directly into the lesions. This could potentially allow some patients to avoid surgery and the negative side effects that can come with intravenous infusions. The implications could extend to other types of precancerous lesions.
“In this trial, we wanted to investigate whether a simple, targeted injection of immunotherapy directly into the lesion could shrink it and retrain the immune system to fight it. If successful, we're not just treating a lesion; we're giving patients back their voice,” says Amit.
AI model may help predict immunotherapy response in some patients with lung cancer
One of the most significant challenges in treating metastatic non-small cell lung cancer is that it is a very diverse disease, and clinicians do not have many reliable methods for predicting how patients will respond to treatment.
Rukhmini Bandyopadhyay, Ph.D., a postdoctoral fellow at UT MD Anderson, and colleagues in the lab of Jia Wu, Ph.D., are harnessing the power of artificial intelligence to improve the ability to make these predictions and better inform key treatment decisions.
Their platform, Path-IO, works by identifying specific features within the tumor microenvironment to understand how tumors and the surrounding tissues are organized. The model then combines this information with available imaging and clinical data to estimate how likely a patient is to benefit from immunotherapy.
Path-IO, as the name indicates, is based on an emerging field known as pathomics. Pathomics tools use AI to visually recognize structures and patterns that humans may miss, but they are grounded in their known roles in immunotherapy response. As a result, the predictions they make are explainable, compared to so-called “black box” AI tools, which is a significant distinction for their potential to reach the clinic.
“There is still some work to do before they’re ready for the clinic, but pathomics platforms like Path-IO will hopefully become extremely valuable tools to help clinicians make faster and more informed decisions for their patients,” says Bandyopadhyay.
Bandyopadhyay will present the results of this study on April 20.
Novel treatment combination shows promising antitumor activity in some advanced tumors
Timothy Yap, M.B.B.S., Ph.D., will present the first clinical data on a novel combination treatment for advanced solid tumors with specific genetic alterations.
Preclinical data showed significant potential for combining the WEE1 inhibitor zedoresertb and the PKMYT1 inhibitor lunresertib for patients with some genetic alterations. In models, the two worked well together because of the way cancer cells respond to each of them as a monotherapy.
This effect pushes cells past their ability to repair their own DNA, inducing a type of cell death.
"Cancers with these specific alterations represent areas of unmet clinical need with no available standard-of-care targeted treatment options,” explains Yap. “If this combination is successful and well tolerated, it could be a significant milestone, especially in ovarian cancers where these alterations are more common.”
Yap will present the full data from the trial on April 19.
Albert Koong, M.D., Ph.D., is chief scientific officer at UT MD Anderson.