Two smokers, exposed to roughly the same amount of tobacco smoke over similar years, can end up with completely different outcomes. One develops lung cancer. The other doesn't. Doctors have always suspected inherited genetics plays some role in that gap, but proving it directly in humans has been nearly impossible, since real people's lives involve too many tangled variables, diet, environment, timing, dose, to isolate genetics as the deciding factor. A study published July 22, 2026, in the journal Nature, led by researchers at the University of Cambridge, found a way around that problem entirely: rerun cancer's earliest evolution hundreds of times over, under identical conditions, and watch what inherited genetics alone actually changes.

Borrowing an idea from evolutionary biology

The study's design draws directly on a famous thought experiment posed by the paleontologist Stephen Jay Gould, who once asked what would happen if you could rewind the tape of evolutionary history and let it play out again from the same starting point. Would life look the same the second time around, or would small differences compound into an entirely different outcome? Sarah Aitken, first author of the study and now an assistant professor of pathology at Yale School of Medicine, who conducted much of the work while at the Cancer Research UK Cambridge Institute, applied that same logic directly to cancer.

"We could remove a lot of that heterogeneity that exists in human populations," Aitken explained, describing the rationale behind replaying tumor evolution under laboratory conditions where every variable except inherited genetics could be held constant.

A carefully controlled experiment across four mouse strains

To pull this off, the research team used four genetically distinct mouse strains specifically bred to carry a range of inherited genetic diversity comparable to what exists across human populations. Every animal received an identical single dose of diethylnitrosamine, a well-studied DNA-damaging carcinogen, at the same developmental stage, using an established liver cancer model. Sex, environment, and carcinogen exposure were all controlled identically across every strain. The only variable left to differ between the mice was the genetic background each strain carried into the experiment.

The team then sequenced the genomes of almost 600 resulting tumors, layered in transcriptomic data showing which genes were actively switched on within each tumor, and examined histopathology slides, essentially reconstructing each tumor's evolutionary path from its original cancer-triggering mutation forward. Separate groups of untreated mice from the same strains were also studied to compare rates of spontaneous tumor formation, giving researchers a baseline against which to measure the carcinogen's specific effects.

What genetic background actually changed about the tumors

The results confirmed that inherited genetic background does far more than simply raise or lower overall cancer risk in the abstract. According to the study, epistatic interactions, meaning interactions between an animal's inherited genetic background and the new mutations a tumor acquires as it develops, directly shaped population-specific patterns in how disease progressed. That included which specific driver mutations a tumor was more likely to select, how frequently whole-genome duplication occurred, a significant event in which a cell's entire genetic content doubles and often accelerates cancer progression, and the broader dynamics of subclonal selection, meaning which competing cell lineages within a developing tumor ultimately won out and came to dominate it.

Put plainly, mice with different inherited genetic backgrounds exposed to the exact same carcinogenic dose developed tumors that evolved along genuinely different molecular paths, not just tumors that appeared with different frequency. That distinction matters enormously for how researchers think about cancer prevention and treatment, since it suggests genetic background isn't merely a dial that turns cancer risk up or down, it actively shapes the specific biological character of whatever tumor eventually forms.

Why this challenges a core assumption in precision oncology

Modern precision oncology largely operates on the assumption that a tumor's driver mutations, genes like KRAS, BRAF, EGFR, or HRAS, are the primary factor determining how a cancer will behave and which targeted therapies are likely to work. Jonathan D. Grinstein, writing for Inside Precision Medicine, summarized the tension this study exposes: two patients sharing the identical driver mutation can nonetheless have meaningfully different diseases, developing and progressing at different speeds and responding differently to identical treatments.

Aitken described the underlying reality driving that inconsistency in blunt terms: "We all have different inherited genes, and we all live different lives. We're exposed to different things." Her team's mouse experiments provide direct, controlled evidence that inherited genetic background is a real, quantifiable part of that variability, not simply background noise obscuring an otherwise consistent biological process.

What this could mean for screening and treatment going forward

Aitken was direct about the practical implications she believes this research points toward. "If genetic background influences both cancer risk and the evolutionary trajectory of tumors, future cancer prevention and screening strategies will need to take into account inherited genetics and population diversity," she said. She extended that reasoning to treatment as well: "Similarly, how people respond to cancer drugs is likely to differ depending on their inherited genetics, and so we may need to tailor our diagnostics and treatments accordingly."

Dr. Sam Godfrey, research information lead at Cancer Research UK, one of the study's funders alongside the Medical Research Council, the European Research Council, and Wellcome, offered a similarly forward-looking but appropriately cautious assessment. "This study gives us a fascinating hint that our inherited genes might have a big influence on the way that cancers develop after DNA damage," Godfrey said. "We still need to see more research before we can understand what this means in humans, but this finding could change our understanding of how cancer starts, and lead to more powerful and precise ways of tackling cancer."

The gap between mouse models and human cancer care

The study's authors are careful to frame their findings as an important mechanistic proof of concept rather than an immediately actionable clinical tool. Mouse strains, however genetically diverse by laboratory standards, do not fully replicate the scale and complexity of human genetic variation, nor the decades-long, messy accumulation of environmental exposures that shape real-world human cancer risk. Translating this finding into concrete changes in how doctors screen for cancer or select treatments will require substantially more research directly examining human populations and their tumors.

Still, the experimental design itself represents a genuine methodological advance. By finding a way to isolate genetic background as a variable, something no observational human study can cleanly achieve, the Cambridge-led team has given cancer researchers a controlled, repeatable model for asking questions about inherited risk that were previously nearly impossible to answer with this level of precision. Whether that translates into new screening guidelines or genetically tailored treatment protocols will depend on how well these mouse-model patterns hold up as researchers begin testing them against real human genomic and clinical data.

---

*Sources cited in this article include the peer-reviewed study "Genetic background sets the trajectory of experimental cancer evolution," published July 22, 2026, in Nature, and reporting from ScienceDaily, the University of Cambridge, EurekAlert, Inside Precision Medicine, Medical Daily, and News-Medical.net. All figures reflect reporting available as of July 30, 2026.*