During the Plenary Session “Early-Onset Cancers,” experts examined the rising incidence of early-onset cancers and the potential contributions of obesity, bacterial exposures, and weaning to this trend.

“The topic of early-onset cancers has really captured the attention of the cancer research community, our patients, and their families,” said Session Chair Andrew T. Chan, MD, MPH, of Massachusetts General Hospital, noting that presentations in the session would examine potential factors behind their increasing rates. “Understanding the why will help us understand how to prevent these cancers in the future.”

The first speaker of the session was Hyuna Sung, PhD, of the American Cancer Society, who argued that, since age is a “moving target,” it is more helpful to examine cancer rates based on when a person was born, meaning their birth cohort, rather than their current age.

She explained that the birth cohort approach better captures risk across a person’s lifetime, reflects early-life exposures that might differ across generations, more fairly compares risk across generations, and reveals generational risk trajectories.

Sung shared data from a birth cohort analysis that found that some cancers, such as colorectal cancer, were becoming more common among younger generations but not older ones. This suggests a birth cohort-specific driver, Sung said.

In contrast, some other cancer types are rising in both younger and older generations. This is indicative of a period effect, which is something that affects all age groups simultaneously, such as changes in diagnostic approaches or disease classification.

Sung argued, however, that most cancers are likely driven by a combination of birth cohort-specific and period-specific factors.

“Even for cancers with a strong birth cohort effect, we cannot fully discount the influence of increasingly sensitive diagnostics and expanding use of these technologies over time. On the flip side, not all incidentally diagnosed cancers are simply overdiagnosis,” she said.

The next presenter was Chan, who dug deeper into lifestyle factors that might be driving early-onset colorectal cancers. Leveraging longitudinal cohort data from the Nurses’ Health Studies I and II and the Health Professionals Follow-Up Study, Chan and colleagues determined that adolescent/early adulthood and/or mid-life obesity was associated with an increased risk of early-onset colorectal cancer.

Additional studies suggested that the contributions of obesity may begin even earlier. “We have some early evidence that even in utero exposure to obesity, adiposity, and an obesogenic environment may be linked to early-onset cancer,” Chan said.

The evidence came from information provided by the cohort participants’ mothers about their own exposures during pregnancy. This revealed higher colorectal cancer risk among participants whose mothers experienced greater weight gain during pregnancy.

Chan also touched on the potential contributions of ultraprocessed foods to early-onset cancer, and, in animal models, discovered that early-life exposure to inflammation could lead to enduring epigenetic changes that may predispose an individual to cancer development.

Building on the association between obesity and colorectal cancer risk, Chan shared that he and colleagues are conducting the PROSPECT trial to evaluate a GLP-1 receptor agonist for colorectal cancer prevention. The trial has enrolled 20 individuals who previously underwent resection of a precancerous adenoma. The participants are being treated with tirzepatide and providing tissue and fluid samples over time, which the researchers hope will help identify epigenetic markers that change upon obesity reversal.

Ludmil Alexandrov, PhD, of University of California, San Diego, discussed the contributions of early-life mutagenesis to the rise of early-onset colorectal cancer. By sequencing nearly 1,000 treatment-naïve colorectal tumors, including 132 early-onset tumors, from around the world, Alexandrov and colleagues found that early-onset tumors were enriched for three mutational signatures. Two of these signatures were associated with colibactin, a genotoxic metabolite produced by certain bacteria.

Despite the presence of colibactin-associated signatures in early-onset tumors, none of the tumors were colonized by colibactin-producing bacteria, suggesting that the mutational signatures developed in response to an earlier exposure to the bacteria. Through analyses of tumor evolution, Alexandrov and colleagues determined that the development of the colibactin signatures was one of the earliest steps in the formation of colorectal cancer. Further, they determined that these signatures were likely generated within the first five years of life.

Additional analyses revealed that, in tumors with the colibactin signature, more than 25% of mutations in the APC gene (a key driver of colorectal cancer) were generated by colibactin.

“Not only did we have the colibactin generating mutational [signatures], but [it was also] generating some of the first driver mutations in colorectal cancer evolution,” Alexandrov said. These results have led him to hypothesize that early-life exposure to microbiome-derived carcinogens, such as colibactin, may be an important factor contributing to the recent increase in early-onset cancer rates.

The final presentation of the session was delivered by Pepper Schedin, PhD, of Oregon Health and Science University, who discussed the biological changes that occur in the postpartum period that might contribute to early-onset breast cancer spread.

Schedin noted that patients diagnosed with breast cancer within 10 years of pregnancy have a higher risk of metastasis than those who have never been pregnant, and the increased risk may be due to weaning-induced involution of the mammary gland and liver.

Involution, or remodeling, of the mammary gland occurs when the mother stops breastfeeding or immediately after pregnancy if the mother does not breastfeed. Involution relies on several wound-healing processes that can also promote tumor progression. These include epithelial cell death (which triggers immune suppression) and loss of the myoepithelial cell barrier (which facilitates shedding of cancer cells from the tumor).

Schedin and colleagues found that the liver, a common site of metastasis in patients with postpartum breast cancer, expands during breastfeeding to support lactation and undergoes involution during weaning. This leads to changes in the liver that may support metastasis, Schedin said.

“Our model is that during weaning-induced involution, the liver has a myriad of changes that create a premetastatic niche,” she said. “For the unfortunate group of young women who may have a tumor, it makes a perfect site for metastasis and could explain the increased risk of liver metastasis that is specifically observed in postpartum patients.”

Schedin also reported that tumor cells that metastasize to the liver readily integrate into the liver environment, as an example of “replacement growth.”

“The hepatocytes are replaced with tumor cells. There is not a massive tumor that forms but rather the tumor seems to use the existing architecture of the liver to establish itself,” she said. Although this feature presents a challenge, Schedin hopes it will also provide therapeutic opportunities for postpartum breast cancers.

The recording of the full session is available on demand for registered Annual Meeting attendees through October 2026 via the virtual meeting platform.

More from the AACR Annual Meeting 2026 »

View a photo gallery of scenes from San Diego, join the conversation on social media using the hashtag #AACR26, and read more coverage in AACR Annual Meeting News and on Cancer Research Catalyst, the official blog of the AACR.