Breast Cancer

Aspirin is one of the most studied repurposed drugs in breast cancer.

It also carries the clearest caution in this space. A large Phase 3 randomised trial in unselected high-risk breast-cancer patients found no benefit from daily aspirin at 300 mg. At the same time, large observational cohorts, mechanistic work, and molecular subgroup data still suggest a real but context-dependent signal — especially around PIK3CA, low-dose long-term use, and metastasis-related platelet biology.

This page holds both sides of that picture together honestly.

Overview

Breast cancer is not one disease.

That matters here more than usual. Aspirin's relevance appears to shift with dose, timing, molecular subtype, and whether the main question is tumour control, recurrence, or metastatic spread.

The biological rationale is strong. Aspirin affects COX-2, platelet TXA2 signalling, PI3K pathway sensitivity, and immune escape during the metastatic window. The clinical evidence is more mixed.

Key human data

Phase 3 trial — the important negative

The most important prospective result is the Alliance 011502 trial, published in the Journal of the American Medical Association in 2024. It randomised 3,020 patients with high-risk, non-metastatic, HER2-negative stage II or III breast cancer to aspirin 300 mg/day or placebo after standard treatment. Read the full trial report

The trial stopped early at interim analysis. Aspirin showed no improvement in invasive disease-free survival and no overall survival benefit. Deaths were numerically higher in the aspirin arm, though not interpreted as proof of harm from aspirin itself. Read the trial summary in The ASCO Post

Aspirin should not be assumed to help all breast-cancer patients at all doses simply because it looks promising in other tumours, in other doses or in observational datasets. Read a plain-language summary of the negative result

Observational evidence — where the signal lives

The negative Phase 3 result does not erase the rest of the literature.

Several large observational studies still suggest a recurrence or mortality signal, especially with low-dose and long-duration use.

• Danish cohort, British Journal of Cancer 2025: Among 20,509 breast-cancer survivors, 20-year cumulative recurrence was 17.8% in low-dose aspirin users versus 22.4% in non-users. The association stayed broadly consistent across ER status, grade, and stage. Read the British Journal of Cancer report

• NHS and NHSII cohort, 2025: Regular post-diagnostic aspirin use was associated with 38% lower breast-cancer-specific mortality and 28% lower total mortality over up to 34 years of follow-up. The association looked stronger with longer duration. It weakened in older patients and higher-stage disease. Read the open-access cohort paper

• Pre-diagnostic use: Long-term aspirin use before diagnosis was linked to downregulation of tumour proliferation pathways in tumour tissue. An aspirin-related gene-expression signature also tracked with better survival in the METABRIC dataset. Read the open-access cohort paper

• Low-dose specificity: Regular low-dose aspirin was associated with a 16% lower breast-cancer incidence in one large analysis. The same signal was not seen with 325 mg aspirin or other NSAIDs. Read the Harvard Health summary

• Meta-analysis, The Oncologist 2024: Pooled hazard ratio for recurrence was 0.89 with a 95% CI of 0.67 to 1.16. That is directionally favourable, but not statistically significant overall. Read the meta-analysis in The Oncologist

• Chinese cohort data, 2025: Additional observational work in Chinese breast-cancer populations also supports a possible mortality benefit with regular aspirin use. Read the Cancer Epidemiology, Biomarkers & Prevention paper

The mortality paradox

One result needs careful handling.

The 2025 Danish cohort found lower recurrence in aspirin users, but also higher all-cause and breast-cancer-specific mortality. The authors argue this likely reflects confounding by indication. In real-world cohorts, aspirin users often have more cardiovascular disease and more baseline comorbidity than non-users. View the Danish cohort record

That does not erase the recurrence signal. It does mean mortality findings from observational studies need more caution than they first appear to.

Molecular subgroup data

PIK3CA mutation

This is the most biologically compelling subgroup in breast cancer.

Preclinical work shows aspirin suppresses PIK3CA-mutant breast-cancer cells more strongly than wild-type cells. The mechanism appears to involve AMPK activation and mTORC1 inhibition, rather than simple COX-2 suppression alone. Read the open-access mechanistic study

In these models, aspirin also increases sensitivity to PI3K inhibitors. That makes combination logic plausible, at least biologically. Read the open-access mechanistic study

Observational metastatic data also point in the same direction. In one study, aspirin users with PIK3CA-mutant breast cancer had longer time to metastasis than non-users, 7.1 years versus 4.7 years. Read the open-access metastatic study

The key limit is obvious. Breast cancer still has no PIK3CA-stratified aspirin RCT. Colorectal cancer has now reached that stage. Breast cancer has not. Read the ALASCCA trial publication

HER2-positive disease

HER2-positive disease stays largely unanswered in trial terms.

Preclinical work from 2017 showed that TXA2 blockade suppressed HER2-driven mammary tumour formation and metastasis in mouse models. Read the open-access mechanistic study

But the Alliance trial enrolled HER2-negative patients only. That means the most important breast-cancer aspirin RCT does not settle the question for HER2-positive disease. View the trial registration

TXA2 and the metastatic window

This is where aspirin's breast-cancer story becomes more interesting than a simple inflammation narrative.

Breast tumours, especially ER-negative, PR-negative, and triple-negative disease, can produce more TXA2 than matched normal tissue. Higher TXA2 tracks with larger tumours and greater metastatic potential. Read the breast-tumour TXA2 study

Breast-cancer cells also express TBXA2R, the thromboxane receptor. That supports migration, invasion, survival, and resistance to oxidative damage in aggressive models. Read the open-access mechanistic study

Platelets add a second TXA2 source. When they coat circulating tumour cells, they help those cells survive in the bloodstream and seed distant organs. Read the open-access platelet metastasis review

The 2025 Nature study by Yang and colleagues sharpened this further. It identified a TXA2 → ARHGEF1 → CD8+ T-cell suppression axis that acts like a systemic immune checkpoint during metastatic spread. Aspirin disrupted that pathway through COX-1 blockade and restored immune control of metastasis in experimental models. Read the University of Cambridge research summary

This may help explain the mismatch between datasets.

Observational studies may be picking up an anti-metastatic effect during a vulnerable circulating-cell phase. The Phase 3 adjuvant trial may have tested aspirin in a clinical setting where that effect was diluted, mistimed, or outweighed by other variables.

This is still a hypothesis, not proof. But it is one of the strongest current explanations for why breast-cancer aspirin data can look genuinely encouraging and genuinely disappointing at the same time.

Clinical positioning

Practical interpretation

The breast-cancer aspirin literature teaches a broader lesson.

Dose matters. Subtype matters. Timing matters.

The negative Alliance trial tested 300 mg aspirin in an unselected high-risk population. That is not the same question as long-term low-dose aspirin in a biologically selected subgroup, or aspirin's effect during the metastatic window.

For patients the most useful questions are:

1. Has the tumour been tested for PIK3CA mutation?

2. What is the tumour subtype, and is platelet/TXA2 biology likely relevant?

3. Given current treatment, GI risk, and bleeding risk, is low-dose aspirin a reasonable discussion with the oncology team?

The answer is not a blanket yes or no.

It depends on tumour biology, comorbidities, treatment context, and how much weight one gives to a strong negative RCT versus suggestive subgroup and observational data. Read the British Journal of Cancer report

References

• Alliance 011502 trial, the Journal of the American Medical Association 2024 — Phase 3, 300 mg/day, no benefit. Read the full trial report

• British Journal of Cancer 2025 — low-dose aspirin and 20-year recurrence findings in a Danish cohort. View the Danish cohort record

• NHS and NHSII cohort, 2025 — post-diagnostic aspirin use and mortality associations. Read the open-access cohort paper

• Preclinical PIK3CA paper, 2016 — aspirin, AMPK, and mTORC1 in breast cancer. Read the open-access mechanistic study

• Metastatic PIK3CA-linked observational study, 2020. Read the open-access metastatic study

• Yang et al., Nature 2025 — TXA2, ARHGEF1, and CD8+ T-cell suppression in metastasis. Read the University of Cambridge research summary

• Low-dose prevention signal overview. Read the Harvard Health summary

• Meta-analysis, The Oncologist 2024 — post-diagnostic aspirin and breast-cancer outcomes. Read the meta-analysis in The Oncologist

Would you like to ask Abbey about the information shared on this page? Would you like to contribute your experience, research or ideas to this page? Perhaps you want to point out something that needs changing?

Feedback Form