A computer model estimates how much chemotherapy-driven aging could cost breast cancer survivors

Chemotherapy kills cancer cells, but it also damages healthy ones. Some of the damaged cells do not die: they stop dividing and linger, releasing inflammatory signals. These senescent cells pile up naturally as we age, and several studies have found more of them in women after chemotherapy for breast cancer. That has raised the worry that the treatment may speed up aging itself, bringing heart disease, diabetes and other age-related illness earlier.

How much might that matter over a whole lifetime? Nobody has followed survivors for 40 years with the right measurements. So a team at Georgetown University and partner centres asked a computer instead. It is important to read everything below as a projection: the numbers come from a simulation, and they are only as good as the assumptions put into it.

What they did

The team started from a long-established breast cancer model, one of a group of models that have been used to inform screening guidelines in the US. It simulates the life of each woman, one by one, from birth to death: whether she develops breast cancer, how it is treated, and when and why she dies.

The new ingredient was a measure of biological age. The researchers used the activity of a gene called p16, a common marker of senescent cells, which rises with age. From published data they built the normal spread of p16 levels at each age, then used clinical measurements of how much p16 goes up after chemotherapy. Each simulated survivor's higher p16 level was translated into an older biological age, and that biological age, not her birthday age, set her risk of dying from causes other than breast cancer.

They ran the whole population twice, much like the two arms of a trial: once with every woman who got breast cancer receiving an anthracycline-based regimen (a stronger, more toxic type of chemotherapy), and once with all of them receiving a regimen without anthracyclines. To check the model, they compared its 8-year predictions with the results of the ABC trials, which compared these two kinds of treatment in real patients.

What they found

• The check against real trials held up. Like the ABC trials, the model found better breast cancer survival with anthracyclines but no difference in overall deaths over 8 years.
• Years of life lost. Survivors given anthracyclines were projected to lose 9.4 years of remaining life (interquartile range, the middle half of results, 7.6 to 11.2) to faster aging. Compared with the gentler regimen, the median difference was 7.6 years.
• Other causes take over sooner. Over time, the risk of dying from something other than breast cancer overtakes the risk from the cancer. Going by birthday age alone, that crossover came 33 years after diagnosis; with anthracycline-related aging it came a median of 16 years earlier. For women diagnosed in their 40s it came 19 years earlier, and the shift was smaller for women diagnosed in their 60s.
• Youngest women had the most to lose. For women treated at 30 to 39, the model projected a median of 9.6 more life-years lost with the stronger regimen than with the gentler one.
• The gap grows slowly. In the first 5 to 10 years, overall survival with the two regimens stayed within 5% of each other. After 20 years, overall death rates were 10% to 25% higher after the stronger regimen in the model.

What to keep in mind

The researchers list several limits. Biological age rested on a single marker, p16, which is not used in routine care. The two scenarios ignore the reasons real women receive one regimen or another, so the results may not apply to particular groups. Hormone therapy, immunotherapy and newer drugs were not modelled. The model assumed baseline p16 has no link with the chance of getting breast cancer, though the evidence on that is mixed. It also assumed the extra aging never recedes after treatment — an assumption that long-term follow-up would need to test. And the input data could not capture how social and environmental stresses change aging from person to person.

The model also predicted lower survival from other causes than the ABC trials saw, probably because it used death rates for the general population, while trial volunteers tend to be healthier.

The team call their results "illustrative", a demonstration of what such a model can do. They state plainly that, because treatment aims to improve breast cancer survival, oncologists should not change treatment decisions because of possible faster aging. In the model, the stronger regimen still improved survival from breast cancer.

Why it matters

Cancer survivors now live long enough for the late costs of treatment to count. This study puts a rough size on one of them and shows where it is likely to weigh most: women treated young, with the harshest drugs. The authors suggest that biological age could become a measured outcome in trials of new cancer drugs, and that knowing who is at risk may help target care after treatment. Before any of that reaches the clinic, they write, studies will have to show that measuring and managing biological age actually improves survivors' lives.