Potential mechanistic basis for aspirin preventing cancer metastases

Yang et al 2025 describe a possible mechanistic basis for reducing metastatic disease burden with aspirin treatment. This post looks at their work based on mice/cell lines and the possible risk-benefit of applying this to humans.

Metastasizing cancer cells are vulnerable to immune attack because they must leave their immunosupressive TME to go to different areas. TXA2 acts on T cells to triggere immunosupressive pathway and inhibition of TXA2 by COX-1 inhibitors free T-cells from this state.

Meta-analyses of large RCTs demonstrate that daily ASA is associated with decreased metastasis (HR 0.64 [0.48-0.84]). Low-dose ASA is associated with lower death rate in non-metastatic cancer patients (HR 0.49 [0.3-0.79]).

ARHGEF1 when deficient led to reduced metastases (mets) when mice were injected with various cancer cell line strains (lung, melanoma). T-cell specific ablation of ARHGEF1 led to lower mets. Other WBC lines with ARHGEF1 ablated did not make meaningful differences in met rates.

Most candidate GPCRs had little effect on T-cell activation and acted in an ARHGEF1 independent manner. However, it was observed that TXA2 acts on a specific GPCR (TXA2-R) through ARHGEF1 to suppresses T-cells. TXA2-R signalling supressed T cell proliferation in the absence of other cell types.

Aspirin is a known TXA2 blocker. Aspirin reduced met frequency in control mice but not in ARHGEF1 deficient mice. Aspirin in the presence of a TXA2 analogue did not reduce metastatic load compared to control in mice.

This paper was a delight to read. Hypothesis driven science with rigorous methods and experiments revealing a mechanistic basis for a possible anti-cancer treatment. Zooming out, aspirin is a relatively harmless drug for the vast majority of patients. The major contraindications for low-dose aspirin are likely hemophilias, children (possibility for Reye’s), allergy or aspirin-exacerbated respiratory disease. Aspirin can increase the risk of bleeds specifically GI and intracranial bleeds. The American Gastroenterolical Association notes from meta-analyses of CVD prevention trials that low-dose aspirin increases the odds of major GI bleeding by 1.59 [1.32-1.91] and 1.34 [1.07-1.7] during a follow up period of 3-10 years source 1 and source 2. For intracranial bleeds, there were 144 events in 57451 people in aspirin group versus 103.5 events in 55815 people. This translates to a 0.066% difference in event rates and given 57451 aspirin treated individuals, this roughly means 38 extra events occurred over 3-10 year follow up. Doing the same for the GI group, this means 100 extra events over the follow up period. Bleeds are serious and this risk needs to be weighed against its potential anti-metastatic probability which is hard to know outside of a RCT measuring this.

The ADD-ASPIRIN trial should help inform if aspirin is useful for metastatic disease prevention. Enrollment criteria is non-metastatic breast, colon, esopahgeous, stomach, or prostate cancer patients who have received standard of care (SoC) therapy. Aspirin will be used as an additive. We can compare overall survival between SOC and SOC + aspirin to understand if there is an effect. Overall survival is a good metric because we think aspirin decreases metastatic load and metastases contribute to most of cancer-related deaths. The observational period needs to be long probably on the order of 5-10 years based on signal from previous meta-analyses. We can also observe adverse effects specifically bleeds between the groups to understand the risk profile of ASA therapy in this context too.

Note that approximately 33% of adults of US adults over 40 years old take low-dose aspirin already per NHANES 2011-2012 data estimates. So, if aspirin’s anti-metatstatic potential is a null result, then we can expect a moderate uptick in bleeds. If it is not a null result, this low-cost intervention could have a good positive impact on patients.

Given that Aspirin works by TXA2 inhibition and appears to be ARHGEF1 dependent, patients with high ARHGEF1 expression in T-cells or those whose tumors are highly immunosuppressive (e.g., low tumor infiltrating lymphocytes on pathology sample, low tumor mutational burden, high PD-L1).