Every oncologist has encountered the puzzle. A patient who smoked two packs a day for 50 years dies of a heart attack at 85 with pristine lungs. Another, a nonsmoker who ran marathons and ate kale, is diagnosed with Stage 4 lung cancer at 42. A woman inherits a BRCA1 mutation that should all but guarantee breast cancer, yet she reaches 90 without ever hearing the word “malignancy.” The centenarian sidesteps the disease entirely, despite the fact that cancer risk rises steadily with every birthday.
Cancer research has historically asked one question: what causes it? The answers have been enormously productive. We know the carcinogens, the oncogenes, the tumor suppressor pathways, the chromosomal rearrangements. But causation is not the whole story. If it were, everyone with a BRCA mutation would get breast cancer, and every longtime smoker would get lung cancer. They do not. Something else is at work.
That something, a growing number of scientists believe, may reside in a class of immune proteins called autoantibodies. And a bold new international project called ATLAS, backed by up to $25 million from Cancer Grand Challenges, is now preparing to find out.
Flipping the Question
The ATLAS acronym stands for Antibody Tracking for Long-term Avoidance and Surveillance. It was one of five research teams selected this year by Cancer Grand Challenges, a global funding initiative co-founded in 2020 by the National Cancer Institute and Cancer Research UK. The total investment across all five teams is $125 million. But ATLAS occupies a unique conceptual space: while its sister teams investigate new drivers of cancer, ATLAS is explicitly devoted to understanding why cancer does not happen.
“The cancer avoidance challenge flips conventional cancer research on its head,” the program notes, “asking why some people do not develop cancer, despite being at high risk, be that due to genetic or environmental predisposition.”
The team will be led by Dr. Paul Bastard of the Imagine Institute at the Necker Hospital for Sick Children in Paris. But the project’s intellectual backbone and its most prominent U.S. figure is Dr. Jean-Laurent Casanova, an HHMI Investigator and member of both the National Academy of Sciences and the National Academy of Medicine. On July 1, 2026, Casanova moved his entire laboratory from Rockefeller University to UT Southwestern Medical Center in Dallas, bringing more than 40 scientists, trainees, and students with him. The move was accompanied by a $5 million Governor’s University Research Initiative grant from the state of Texas. Casanova also recently received the 2026 Mechthild Esser Nemmers Prize in Medical Science and the 2025 Novo Nordisk Prize, two of the highest honors in biomedical research.
Autoantibodies: The Body’s Double-Edged Sword
To understand what ATLAS is hunting, you first have to understand a peculiar feature of the human immune system. Humans carry billions of different antibodies, proteins produced by B cells that can bind to virtually any molecular shape. Most of these antibodies are designed to recognize foreign invaders: viruses, bacteria, fungi, toxins. But a fraction of them, by random chance or by breakdown in immune tolerance, recognize the body’s own molecules. These are autoantibodies.
In autoimmune diseases such as lupus or rheumatoid arthritis, autoantibodies are the villains. They attack healthy tissue and cause chronic inflammation and organ damage. But Casanova and his colleagues have spent the last several years demonstrating that the story is more nuanced. During the COVID-19 pandemic, his team discovered that roughly 10 to 15 percent of people who developed life-threatening COVID-19 carried autoantibodies that neutralized type I interferons, the immune system’s first line of antiviral defense [1]. These autoantibodies were present before infection. They did not cause disease in the usual sense, but they disabled a critical protective mechanism, turning what should have been a mild illness into a deadly one.
This finding upended the conventional understanding of autoantibodies. It showed that they could act as “hidden” risk factors, silently undermining specific immune functions without producing the obvious symptoms of autoimmune disease. And it raised a provocative question: if autoantibodies can disable the immune response to a virus, could they also disable the immune response to cancer?
The Hypothesis
The ATLAS team’s core hypothesis is that cancer, like a viral infection, requires the immune system’s permission. The body’s immune surveillance system constantly patrols for abnormal cells, and most early tumors are eliminated before they become clinically detectable. But if certain autoantibodies neutralize key components of that surveillance system, tumors may slip through. Conversely, some autoantibodies might actually enhance antitumor immunity, offering protection that conventional risk models cannot explain.
“There is ample evidence that the immune system plays a significant role in cancer,” Casanova said. “Our job will be to determine if there are auto-Abs that could offer protection against cancer or drive cancer outcomes, knowledge that could have a profound impact on cancer risk assessment and treatment.”
The team laid out this framework in a commentary published in Cell [2], describing a systematic plan to profile the antibody repertoires of unique human populations and build the first comprehensive Cancer Antibody Atlas.
The Study Cohorts
To separate protective autoantibodies from harmful ones, ATLAS will study several carefully chosen groups. The first are heavy smokers who have somehow avoided lung cancer. These “super-avoiders” represent a direct challenge to conventional risk assessment. If autoantibodies explain their protection, identifying those antibodies could point toward new preventive therapies.
The second group are centenarians and other very old individuals who have remained cancer-free. Cancer risk increases dramatically with age, so anyone who reaches 100 without a cancer diagnosis has likely benefited from some form of biological protection. Studying their immune profiles may reveal autoantibodies that become more effective with age, rather than less.
The third group are identical and fraternal twins in which one twin has developed cancer and the other has not. Because identical twins share nearly all of their DNA, any immune difference between them is likely environmental or epigenetic. If a protective autoantibody appears in the cancer-free twin but not the affected twin, that represents a powerful signal.
Finally, the team will study cancer patients before, during, and after immunotherapy. Immunotherapy works by unleashing the immune system against tumors, but it succeeds in only a fraction of patients. By tracking autoantibody profiles across treatment, the team hopes to identify which autoantibodies predict a good response and which predict resistance.
Building the Atlas
The technical challenge is enormous. Each individual carries a unique repertoire of antibodies built up over a lifetime of exposures. Profiling these repertoires at proteome scale, across tens of thousands of individuals, requires platforms such as protein microarrays and phage immunoprecipitation sequencing (PhIP-Seq). The ATLAS team includes specialists from eight institutions across six countries, including CDI Labs, whose Chief Scientific Officer Dr. Tyler Hulett has been developing antibody profiling methods since his doctoral work.
“Everyone has autoantibodies; they are stable for decades and unique even between identical twins,” Hulett noted. “I believe this team will demonstrate that antibodies are not just protective against cancer, but also create immune-disrupting mistakes which allow tumors to grow. Understanding both sides of this equation will transform the way cancer is understood and reveal new treatments.”
The Larger Vision
If ATLAS succeeds, the implications go far beyond basic biology. Protective autoantibodies could be manufactured and administered to people at high risk of cancer, much like monoclonal antibodies are used for infectious diseases. Harmful autoantibodies could be removed through plasmapheresis or targeted with drugs. And autoantibody profiles could become the basis for blood tests that flag elevated cancer risk years before any tumor is detectable.
None of this is guaranteed. The project faces formidable hurdles, including the sheer diversity of the antibody repertoire and the difficulty of proving causation rather than correlation. But the very act of asking the question represents a shift in how the research community thinks about cancer.
For decades, the field has focused on what goes wrong inside a cell to make it malignant. ATLAS asks a different question: what goes right inside a person to keep that cell in check? It is a question born of a simple, stubborn observation that the data have never been able to explain. Why do some people stay healthy when they should not? The answer, if it exists, is written in the antibodies they carry. ATLAS intends to read that code.
References
1. Bastard P, Rosen LB, Zhang Q, et al. Autoantibodies against type I IFNs in patients with life-threatening COVID-19. Science. 2020;370(6515):eabd4585. doi:10.1126/science.abd4585
2. Bastard P, Casanova JL, Hulett T, et al. The ATLAS Project: Antibody Tracking for Long-term Avoidance and Surveillance. Cell. 2026;S0092-8674(26)00708-7. doi:10.1016/j.cell.2026.06.021
3. Cancer Grand Challenges. ATLAS Team. https://www.cancergrandchallenges.org/atlas. Accessed July 24, 2026.
4. UT Southwestern Medical Center. Autoantibodies could play key role in cancer progression and prevention. July 23, 2026. https://www.utsouthwestern.edu/newsroom/articles/year-2026/july-autoantibodies-cancer-progression-prevention.html
5. UT Southwestern Medical Center. Renowned researcher joining UTSW is among recipients of up to $25 million Cancer Grand Challenges award. March 5, 2026. https://www.utsouthwestern.edu/newsroom/articles/year-2026/march-cancer-grand-challenges-award.html