A year of metformin lowered an inflammation signal in the colon of cancer survivors; aspirin did not

Metformin, the most widely prescribed pill for type 2 diabetes, is also one of the most discussed candidates for slowing aspects of aging, partly because it seems to calm low-grade inflammation. Aspirin, another cheap old drug, is known to lower the risk of colorectal cancer when taken for many years. Both are thought to act on a protein called NF-κB, a switch inside cells that turns on inflammation genes and that is linked to cancer growth. A European team tested both drugs, alone and together, in people who had already had colorectal cancer, and looked directly at the lining of their colon.

The question matters because colorectal cancer often comes back: the paper cites a five-year recurrence risk of roughly 30% to 40% after surgery for stage I to III disease, and new cases are rising, driven mainly by aging populations.

What they did

The ASAMET trial recruited adults aged 18 to 80 whose colorectal cancer had been completely removed within the previous two years. Fifty-one people (42 in Italy, 8 in Slovenia and 1 in Austria) were randomly assigned to one of four groups for 12 months:

• placebo,
• aspirin, 100 mg once a day,
• metformin, 850 mg twice a day,
• both drugs.

The trial was double-blind: neither the patients nor the doctors knew who took what, and every patient swallowed two sets of pills (real or dummy) so that the groups looked the same. Average ages in the groups ranged from about 63 to 68.

Before and after the 12 months, doctors took small tissue samples of normal-looking colon lining during a colonoscopy, from both the right and the left side of the colon. In the lab, the samples were stained to count how many cells showed NF-κB (the main outcome), Ki67 (a marker of cells that are dividing) and pS6K (a sign that the growth pathway mTOR is active).

What they found

• Metformin lowered NF-κB. Compared with placebo, the share of cells staining for NF-κB fell by about 9 percentage points more in the metformin group (P = 0.04). When the two groups taking metformin were combined, the reduction was about 7 points compared with those who did not take it.
• Aspirin did not. Aspirin alone made no clear difference to NF-κB, and the combination showed a similar-sized drop to metformin alone that did not reach statistical significance.
• No overall change in cell division. Ki67 did not change overall with any treatment. But in the right side of the colon, the combination raised Ki67 by about 12 points compared with placebo, which the authors call unexpected.
• No sign of the mTOR pathway changing. No treatment changed pS6K.
• Side effects. Side effects were manageable and in line with what is known for both drugs, and there were no serious adverse events.

What to keep in mind

• Much smaller than planned. The trial aimed for 160 patients but enrolled 51 because of the COVID-19 pandemic and funding problems, and closed early. Only 31 people had both biopsies and were analysed. The authors say all results should be read as "hypothesis-generating".
• A marker, not cancer. The trial measured proteins in tissue samples, not whether cancer came back or how long people lived. Follow-up was too short for that.
• Sampling. A few small biopsies at two time points may not capture the whole colon, and the quality of the tissue varied between centres.
• One region. All centres were in Europe, so the results may not apply everywhere.
• The aspirin finding is narrow. The authors stress that it does not contradict the strong evidence that long-term aspirin lowers colorectal cancer risk; aspirin may work through other routes, for example through blood platelets and the immune system, which these markers don't capture. The rise in cell division with the combination is unexplained.
• Prescription drugs. Metformin is a prescription medicine, and aspirin can cause bleeding. Neither should be started for prevention without a doctor.

Why it matters

This is, according to the authors, the first placebo-controlled trial to look at what metformin and aspirin together do to the normal colon lining of cancer survivors. The main result, that metformin damped an inflammation switch in healthy tissue over a year, fits the idea that the drug can calm inflammation independently of blood sugar, one of the reasons it is studied as an anti-aging drug. The team suggests that a larger phase III trial of metformin for preventing colorectal cancer could be worthwhile. Whether the drop in NF-κB translates into fewer cancers is not yet known.