Dozens of clinicians and scientists from across Memorial Sloan Kettering Cancer Center (MSK) presented the latest advances in cancer research at the 2026 American Association for Cancer Research (AACR) Annual Meeting , held April 17 to 22 in San Diego.
MSK’s presence spanned the entire event — from the opening plenary session, to major symposia, to poster presentations.
MSK researchers presented on innovative therapeutic advances — including clinical trials showing promise for people with early-stage HER2-driven rectal cancer and KRAS-driven pancreatic cancer — and reported on the encouraging results from a phase 1 study of a therapeutic mRNA cancer vaccine to treat pancreatic cancer .
Scientists from MSK also presented on emerging discoveries in immuno-oncology, computational oncology, understanding cancer “ecosystems,” overcoming treatment resistance, and more.
Here are some of the research highlights from this year’s meeting:
A team of researchers from MSK and Columbia University has engineered CAR T cells to selectively eliminate cells that express the surface protein uPAR, which plays a key role in wound healing and tissue repair and reshaping.
In cancer, uPAR is persistently overexpressed in both tumor cells and cancer-supporting cells in the tumor microenvironment.
In preclinical models, CAR T cells engineered to target uPAR showed promising results. The new approach shrank several types of solid tumor in mouse models of lung, pancreatic, and ovarian cancers — and even cleared metastases in some experiments.
CAR T cell therapies have primarily been used to treat blood cancers, but the new approach offers promise for the treatment of solid tumors. That’s because they attack not only a consistent target across a range of tumor types but also the protective network of fibrotic tissue and immune-suppressive cells around a tumor.
Moreover, despite uPAR being present on some normal immune cells, the treatment did not cause sustained depletion of these cells in mice, suggesting a favorable safety profile.
The team’s research was presented at AACR by Zeda Zhang, PhD , a postdoctoral researcher in the lab of co-senior study author Scott Lowe, PhD, Chair of MSK’s Cancer Biology and Genetics Program and a Howard Hughes Medical Institute Investigator.
The research is illustrative of efforts supported by the Marie-Josée and Henry R. Kravis Cancer Ecosystems Project at MSK, for which Dr. Lowe serves as scientific director. The initiative is focused on understanding cancer not simply as a genetic disease but as an interconnected ecosystem of cells, tissues, and signaling networks. By framing cancer in this way, the program aims to enable the development of next-generation therapies that target not only tumor cells, but also the surrounding niche that supports their growth and progression.
Learn more: Target Chronic Inflammation in Cancer and Fibrosis With Engineered Immune Cells (AACR Program) and related Cell paper .
Pancreatic cancers are notorious for their therapy-resistant tumor environment — the non-cancer structural tissue and immune cells surrounding cancer cells — but how this environment develops and functions is poorly understood. In a major symposium, MSK computational biologist Dana Pe’er, PhD presented recent work showing how certain cancer cell types drive their neighbors to participate in a self-reinforcing ecosystem that sustains the tumor.
Working with MSK cancer biologist Tuomas Tammela, PhD , the teams made a surprising discovery in mice: when they killed off a rare type of cancer cell called ’basal’ cells, the entire tumor environment collapsed.
To make this discovery, the team turned to sophisticated computational tools they developed. Scientists can now map gene expression across tumors using spatial transcriptomics, but comparing cellular neighborhoods has been a major challenge. The Pe’er Lab overcame this using an algorithm they developed — called Wasserstein Wormhole — to compare neighborhoods and identify spatial trends. They discovered that basal cells create a neighborhood that attracts myeloid immune cells that protect the tumor. But, when the basal cells are removed, the immune system is better able to mount an attack against the tumor.
In a separate collaboration with MSK cancer biologist Scott Lowe, PhD , the Pe’er lab studied early pancreatic tumor formation in mice. The researchers identified highly adaptable cancer cells — called progenitor cells — that share similarities with basal cells. These cells activate genes that normally suppress tumor growth, but when these protective genes fail, the cells turn cancerous. This suggests that p53 — a protective gene mutated in about half of all cancers — normally acts as a brake on the excessive cell flexibility that can develop after tissue injury.
Using a new computational method, the team tracked how the cellular neighborhoods around progenitor cells change over time, finding they increasingly recruit fibroblasts and myeloid immune cells that suppress the body’s anti-tumor response.
Dr. Pe’er also showed how basic data processing can be improved using segger, a new AI-driven approach that better determines which gene transcripts belong to each cell in imaging data.
These efforts, part of the Marie-Josée and Henry R. Kravis Cancer Ecosystems Project at MSK, show how combining sophisticated mouse models, spatial profiling, and novel data analysis can provide powerful insights into how tumor ecosystems develop and function.
Learn more: Spatial and Single-Cell Omics at Scale: Technologies for Decoding Cancer Ecosystems (AACR Program) and related papers in Cell and on bioRxiv .
The genes you inherit from your parents don’t just affect your chances of getting cancer: They can also influence how a tumor grows and changes over time.
One example that illustrates this concept is mutations in BRCA1 and BRCA2 : Some inherited variations on these genes increase the risk of cancer, but they also make tumors more vulnerable to certain treatments. Beyond a few well-studied mutations like these, however, scientists don’t yet have a good understanding of how inherited gene variants affect tumor development or the response to treatment.
In a major symposium titled “Genetic and Environmental Determinants of Cancer,” MSK computational oncologist Jian Carrot-Zhang, PhD , presented data from a large-scale study that explored how inherited gene variants influence which genetic mutations arise in tumors. The research combined genetic, tumor, and patient data from diverse populations with a range of cancer types — it showed that the genes you’re born with can influence which mutations a tumor acquires as it grows. The study also uncovered some previously unknown connections between biological pathways.
“We found examples where genetic differences between populations may help explain why some groups are more affected by certain mutations in the tumors than others,” says Dr. Carrot-Zhang, whose lab studies how genetic variation affects cancer outcomes. “Since many of the tumor genetic changes we studied are already used to guide treatment decisions, understanding how inherited genetics influence these changes could open new doors for more personalized cancer care.”
The immune system uses specialized cells called CD8+ T cells to seek out and destroy damaged cells, including cancer cells. But when these T cells are constantly exposed to cancer, they become worn out — a condition scientists call “T cell exhaustion.” Cancer immunotherapy drugs work in part by helping the immune system to overcome this exhaustion, but their effectiveness is often limited. Therefore, scientists are looking for new ways to keep these T cells working.
In a major symposium on RNA biology in cancer, MSK physician-scientist Omar Abdel-Wahab, MD , Chair of the Molecular Pharmacology Program, presented new research on the underlying biology of T cell exhaustion. The research concentrated on RNA splicing, the focus of Dr. Abdel-Wahab’s lab. His team set out to determine whether exhausted T cells inside tumors have different versions of RNA than what is found in healthy, active T cells. When they analyzed T cells from patients with melanoma, they found many RNA forms that are uniquely expressed in exhausted T cells.
The researchers then changed the way the RNA was manufactured, or spliced, to express therapeutic proteins that could combat T cell exhaustion.
“When we did this we were able to augment the anti-tumor activity of these T cells in melanoma,” Dr. Abdel-Wahab says. “This approach could eventually help us to develop more effective and precise cancer treatments that work with the body’s natural immune processes.”
Learn more: RNA Processing Regulation of Anti-Tumor Immune Responses (AACR Program)
Antibody-drug conjugates (ADCs) have transformed cancer treatment in recent years. They work by attaching to specific targets on cancer cells and releasing highly potent drugs inside tumors. For example, an ADC called trastuzumab deruxtecan (T-DXd, also known as Enhertu®), targets HER2, a protein found on the surface of many cancer cells.
While T-DXd can be highly effective at first, many people develop resistance to this treatment. Furthermore, due to side effects similar to those from chemotherapy, many doctors and scientists have debated just how “targeted” ADCs really are.
MSK physician-scientist Sarat Chandarlapaty, MD, PhD , presented research clarifying why T-DXd stops working. He and physician-scientist Joshua Drago, MD, MS , studied samples from more than 100 patients whose cancers stopped responding to T-DXd, to understand what changed in the tumors. They learned that in about half of patients’ tumors, HER2 levels went down, sometimes disappearing completely. In other patients, certain gene mutations changed the shape of the HER2 protein so that T-DXd could no longer bind to it.
To overcome this resistance, the researchers tried combining T-DXd with dato-DXd, another ADC that hits a different cancer target, called TROP2. In cell and mouse models, this combination approach was more effective than either treatment alone, even when both treatments were given at very low doses. This approach could sidestep resistance due to target loss and reduce ADC side effects, the researchers say, but the findings need to be confirmed in a clinical trial before they are adopted in practice.
This finding is also significant because it helped confirm that T-DXd is, in fact, binding to HER2 on cancer cells. There are several other approved cancer therapies that also target HER2, so the research has implications for determining which treatments might be more effective when used after T-DXd.
Learn more: Resistance to Anti-HER2 ADCs (AACR Program) and related Cancer Discovery paper .
Most healthy adults have cells harboring potentially cancer-causing mutations in their pancreas, yet only a fraction will ever develop pancreatic cancer. MSK cancer biologist Mara Sherman, PhD , presented on her lab’s work investigating how protective mechanisms in healthy tissue help keep tumor formation in check.
Connective tissue (mesenchyme) in a healthy pancreas produces a signaling molecule called KITL, which plays an important role in maintaining the pancreas’ normal tissue structure and organization. KITL limits tissue plasticity — the ability of cells to change their identity or state. Importantly, KITL is lost during the early stages of tumor formation, and this loss allows cells to become more plastic and helps facilitate cancer progression.
Dr. Sherman and her team demonstrated that when they removed KITL from pancreatic connective tissue in mice, tumors grew faster and the animals didn’t live as long. The results occurred across different contexts: in healthy tissue, after injury, and during cancer development.
Understanding how normal tissue architecture protects against tumor formation could inform new strategies for cancer prevention and early intervention, Dr. Sherman notes.
Learn more: Identifying Stromal Barriers to Pancreatic Tumorigenesis (AACR Program) and related Cancer Discovery paper.
The Epstein-Barr virus (EBV) infects most people worldwide, yet only a small subset of people develop serious complications that can include autoimmune diseases, cancer, or neurological disorders.
Research led by MSK, Baylor University, and AstraZeneca developed a sophisticated data-mining approach to understand why some people experience persistent EBV infection while others don’t. MSK computational biologist Caleb Lareau, PhD , presented the team’s work at AACR.
The researchers discovered they could measure persistent EBV DNA levels by analyzing whole genome sequencing data from existing datasets. Using data from nearly 736,000 participants in the UK Biobank and All of Us research programs, they found consistent links between EBV DNA levels in blood and various diseases, including respiratory, autoimmune, neurological, and cardiovascular conditions.
Through genome-wide association studies, the researchers identified genetic factors that influence EBV persistence. They identified 22 genetic variants linked to both higher active levels of EBV and chronic disease risk — and people with these genetic variants might be more likely to develop chronic EBV-related ailments.
Importantly, the framework used to analyze viral DNA persistence in large genomic datasets could be applied to the study of other viruses that affect human health, potentially revealing new targets for preventing virus-related chronic diseases, including cancer.
Learn more: Population Scale Sequencing Resolves the Determinants and Consequences of Persistence Epstein Barr Virus Infection (AACR Program) and related Nature paper.