# 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. {% hint style="warning" %} The key clinical caution is simple. In unselected high-risk breast cancer, **300 mg aspirin daily did not improve outcomes** in a Phase 3 trial. {% endhint %} ### 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](https://jamanetwork.com/journals/jama/fullarticle/2818110) 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](https://ascopost.com/news/june-2024/aspirin-as-adjuvant-therapy-for-breast-cancer/) 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](https://www.obgproject.com/2024/05/04/rct-results-does-daily-aspirin-reduce-the-risk-of-breast-cancer-recurrence-or-death/) #### 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](https://www.nature.com/articles/s41416-025-03112-3) * **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](https://pmc.ncbi.nlm.nih.gov/articles/PMC12216965/) * **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](https://pmc.ncbi.nlm.nih.gov/articles/PMC12216965/) * **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](https://www.health.harvard.edu/blog/aspirin-and-breast-cancer-risk-how-a-wonder-drug-may-become-more-wonderful-2020102321225) * **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](https://academic.oup.com/oncolo/article/29/1/e1/7207782) * **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](https://aacrjournals.org/cebp/article/34/12/2144/767331/Aspirin-Use-and-Risk-of-Breast-Cancer-Recurrence) #### 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](https://pubmed.ncbi.nlm.nih.gov/40691279/) 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](https://pmc.ncbi.nlm.nih.gov/articles/PMC5290090/) In these models, aspirin also increases sensitivity to **PI3K inhibitors**. That makes combination logic plausible, at least biologically. [Read the open-access mechanistic study](https://pmc.ncbi.nlm.nih.gov/articles/PMC5290090/) 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](https://pmc.ncbi.nlm.nih.gov/articles/PMC7181475/) 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](https://pubmed.ncbi.nlm.nih.gov/40961426/) #### 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](https://pmc.ncbi.nlm.nih.gov/articles/PMC5290090/) 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](https://clinicaltrials.gov/study/NCT02927249) ### 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](https://pubmed.ncbi.nlm.nih.gov/27800646/) 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](https://pmc.ncbi.nlm.nih.gov/articles/PMC5290090/) 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](https://pmc.ncbi.nlm.nih.gov/articles/PMC7617601/) 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](https://www.cam.ac.uk/research/news/scientists-discover-how-aspirin-could-prevent-some-cancers-from-spreading) 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 | Setting | Evidence | Position | | ---------------------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------ | | **Unselected adjuvant, 300 mg/day** | Phase 3 randomised trial showed no benefit. [Read the full trial report](https://jamanetwork.com/journals/jama/fullarticle/2818110) | Not supported | | **Low-dose long-term, 75 to 100 mg/day** | Observational recurrence signal. [Read the British Journal of Cancer report](https://www.nature.com/articles/s41416-025-03112-3) | Investigational | | **PIK3CA-mutant disease** | Strong preclinical rationale and supportive observational data. [Read the open-access mechanistic study](https://pmc.ncbi.nlm.nih.gov/articles/PMC5290090/) | Biologically plausible, but unproven | | **Triple-negative or TXA2-high disease** | Preclinical and immune-metastatic mechanism support. [Read the University of Cambridge research summary](https://www.cam.ac.uk/research/news/scientists-discover-how-aspirin-could-prevent-some-cancers-from-spreading) | Future trial candidate | | **Prevention before diagnosis** | Epidemiological incidence signal only. [Read the Harvard Health summary](https://www.health.harvard.edu/blog/aspirin-and-breast-cancer-risk-how-a-wonder-drug-may-become-more-wonderful-2020102321225) | Not treatment evidence | ### 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](https://www.nature.com/articles/s41416-025-03112-3) ### References * Alliance 011502 trial, the *Journal of the American Medical Association* 2024 — Phase 3, 300 mg/day, no benefit. [Read the full trial report](https://jamanetwork.com/journals/jama/fullarticle/2818110) * *British Journal of Cancer* 2025 — low-dose aspirin and 20-year recurrence findings in a Danish cohort. [View the Danish cohort record](https://pubmed.ncbi.nlm.nih.gov/40691279/) * NHS and NHSII cohort, 2025 — post-diagnostic aspirin use and mortality associations. [Read the open-access cohort paper](https://pmc.ncbi.nlm.nih.gov/articles/PMC12216965/) * Preclinical PIK3CA paper, 2016 — aspirin, AMPK, and mTORC1 in breast cancer. [Read the open-access mechanistic study](https://pmc.ncbi.nlm.nih.gov/articles/PMC5290090/) * Metastatic PIK3CA-linked observational study, 2020. [Read the open-access metastatic study](https://pmc.ncbi.nlm.nih.gov/articles/PMC7181475/) * Yang et al., *Nature* 2025 — TXA2, ARHGEF1, and CD8+ T-cell suppression in metastasis. [Read the University of Cambridge research summary](https://www.cam.ac.uk/research/news/scientists-discover-how-aspirin-could-prevent-some-cancers-from-spreading) * Low-dose prevention signal overview. [Read the Harvard Health summary](https://www.health.harvard.edu/blog/aspirin-and-breast-cancer-risk-how-a-wonder-drug-may-become-more-wonderful-2020102321225) * Meta-analysis, *The Oncologist* 2024 — post-diagnostic aspirin and breast-cancer outcomes. [Read the meta-analysis in The Oncologist](https://academic.oup.com/oncolo/article/29/1/e1/7207782)

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

{% hint style="warning" %} This information is for education only. It is not medical advice, diagnosis, or treatment. Please speak with a qualified clinician before making changes to care, medication, or supplement use. {% endhint %} {% hint style="info" %} © 2026 Abbey Mitchell. All rights reserved. Please share by URL rather than copying page text. {% endhint %} --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://myhealingcommunity.gitbook.io/myhealingcommunity-docs/off-label-drugs-for-cancer/aspirin-in-oncology/aspirin-evidence-by-cancer-type/breast-cancer.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.