Most smokers never get lung cancer. Some people who’ve never touched a cigarette do. That mismatch between exposure and outcome has puzzled cancer researchers for decades, and it’s usually chalked up to some vague combination of “genetics and bad luck.” A new study published in Nature, led by researchers at the University of Cambridge working with the University of Edinburgh and collaborators across Europe and the US, offers the first real, direct evidence for what that genetic influence actually looks like, and it’s more specific than anyone expected.

The problem with studying this in people

Cancer starts when DNA accumulates mutations, errors that let cells grow uncontrollably and ignore the signals that would normally tell a damaged cell to die. Things like cigarette smoke and UV light drive up how much DNA damage a person accumulates, while inherited genetic variation can influence how many of those mutations actually stick. But testing that idea directly in humans is nearly impossible: no two people share an identical lifestyle, environment, and lifetime exposure history, so it’s very hard to isolate genetics as the variable that matters.

A cleaner experiment

To get around that, the team bred four different strains of mice with varying natural susceptibility to liver cancer, spanning roughly the same range of genetic diversity you’d see across human populations. Every mouse then received an identical dose of diethylnitrosamine, a liver carcinogen found in tobacco smoke and some processed foods, at exactly the same age. Because the exposure, timing, and dose were identical across all four strains, any differences that showed up afterward could be attributed to genetic background rather than lifestyle or environment, the variable that’s impossible to control for in human studies.

The researchers then sequenced the genomes of nearly 600 resulting tumors, along with examining untreated mice from each strain to track spontaneous tumor formation, and used that data to reconstruct how each individual tumor evolved from its original cancer-triggering mutation.

Same starting point, different roads

Across all four mouse strains, tumors nearly always picked up a driver mutation that activated the same core cancer-promoting signaling cascade, known as the MAPK pathway, a chain of molecular signals that governs cell growth and differentiation and shows up in a huge range of cancer types.

But that’s where the similarity ended. Depending on which strain a tumor arose in, the specific mutation driving that MAPK activation differed, and that choice of mutation had knock-on effects: it altered which other cancer-associated signaling pathways got switched on, and it strongly influenced whether the tumor’s entire chromosome set ended up doubling, an event called whole-genome duplication that’s associated with more aggressive cancer behavior.

In other words, inherited genetic background didn’t just change whether cancer developed. It shaped the entire evolutionary trajectory the tumor took to get there, funneling the same initial DNA damage down genuinely different molecular paths depending on the animal’s genetic makeup.

What this could mean for cancer prevention and treatment

Because inherited genetics shaped both the mutations tumors acquired and the biological pathways they ended up relying on, the researchers argue this has real implications beyond basic biology. If your genetic background steers which molecular route a tumor takes, it’s plausible that it also shapes how that tumor would respond to specific cancer drugs, many of which are designed to target particular pathways like MAPK signaling.

That points toward cancer prevention, screening, and treatment strategies that account more explicitly for a person’s inherited genetics and population-level genetic diversity, rather than treating cancer risk and cancer treatment response as largely uniform across a given environmental exposure.

The researchers are careful to note that this was done in mice, and translating the findings to humans will require further work. But it’s a rare case of directly proving, rather than just statistically suggesting, that the same environmental insult can be steered down meaningfully different cancer-causing paths purely by the genetic hand an individual was dealt.

Source: S.J. Aitken et al., “Genetic background sets the trajectory of experimental cancer evolution,” Nature, published 22 July 2026. Read the full paper: https://www.nature.com/articles/s41586-026-10821-z