High-Dose Omega-3 Linked to Longer Survival After Bowel Cancer Liver Surgery, Phase 3 Trial Finds

Patients who took high-dose omega-3 capsules before and after surgery to remove bowel cancer that had spread to their liver lived substantially longer than patients given a placebo, according to results from a phase 3 trial presented at the European Society for Medical Oncology (ESMO) Congress 2026 in Madrid. The finding is one of the most striking to emerge from cancer research this year — and also one of the most puzzling, because the same trial failed to meet its main goal.

The EMT2 trial, run across 13 specialist liver surgery centres in the United Kingdom, enrolled 418 patients who were about to undergo liver resection — surgery to cut out bowel cancer that had spread to the liver — with the aim of curing them. Half were randomly assigned to take 4 grams a day of purified eicosapentaenoic acid (EPA), an omega-3 fatty acid supplied as the prescription drug icosapent ethyl, while the other half took identical-looking mineral oil placebo capsules. Treatment began at least a day before surgery and continued for between two and four years, even if the cancer came back.

After five years, an estimated 62.0 per cent of patients in the EPA group were still alive, compared with 45.1 per cent of those in the placebo group, according to the trial results. That translates into a 32 per cent lower hazard of death, with a hazard ratio of 0.68 and a 95 per cent confidence interval of 0.48 to 0.96. The result was statistically significant, at p = 0.03. Median overall survival had not yet been reached in the EPA group, while it stood at 59.0 months in the placebo group.

Yet the trial’s primary endpoint — progression-free survival, the length of time patients lived without their cancer growing or returning — showed almost no difference at all. Median progression-free survival was 13.5 months with EPA and 12.9 months with placebo, a hazard ratio of 0.89 that was not statistically significant (p = 0.33). Five-year progression-free survival was 23.6 per cent with EPA and 24.5 per cent with placebo — effectively identical.

That contradiction sits at the heart of the story. A treatment that did not slow the cancer’s return nevertheless appeared to help patients live longer. The trial’s lead investigator and the independent experts who reviewed the data at ESMO all agree on what that means in practice: the result is intriguing, potentially important — and not yet enough to change how patients are treated.

How the trial was designed

EMT2 — short for the EPA for Metastasis Trial 2 — was a randomised, double-blind, placebo-controlled phase 3 trial, the most rigorous design available for testing a medical treatment. It was investigator-initiated and sponsored by the University of Leeds, funded by the charity Yorkshire Cancer Research, and registered as NCT03428477. The drug company Amarin supplied the EPA and placebo capsules but did not run the trial.

Patients were enrolled between 2018 and 2023 at 13 hepatobiliary surgery centres in England and Wales. Their median age was 64. All were scheduled for liver resection with curative intent — surgery aimed at removing every visible trace of cancer that had spread from the bowel to the liver. They were randomised one-to-one: 209 patients received EPA and 209 received placebo.

The dose was 4 grams a day of highly purified EPA ethyl ester, the same prescription formulation and dose used in the landmark REDUCE-IT cardiovascular outcomes trial, according to Amarin. Treatment started at least one day before the operation and was planned to continue for a minimum of two years and a maximum of four, including after cancer progression. Other cancer treatments were broadly comparable between the two groups, with about 20 per cent of patients in each arm receiving adjuvant chemotherapy after surgery.

The trial was presented as a late-breaking abstract, LBA30, in a proffered paper session at the ESMO Congress in Madrid on Sunday 25 October 2026. Its chief investigator, Professor Mark Hull of the Leeds Institute of Medical Research at the University of Leeds, has studied EPA in bowel cancer for more than a decade.

Why liver metastases matter so much

To understand why the result drew such attention, it helps to understand the disease. Around half of all people diagnosed with colorectal cancer — cancer of the bowel — develop metastases in the liver at some point in the course of their illness. The liver is the first organ that blood draining from the bowel reaches, which is why it is the most common site for bowel cancer to spread.

Surgery to remove those liver metastases is the only treatment that offers a realistic chance of cure or long-term survival. With curative-intent liver resection, five-year survival is roughly 40 to 50 per cent. But the cancer comes back in 50 to 70 per cent of patients — in some series up to 75 per cent — most often within the first two years, and most commonly in the liver itself, followed by the lungs.

That pattern of frequent recurrence is precisely why EMT2’s design mattered. The trial was built on laboratory and observational evidence suggesting EPA might have activity against colorectal cancer, including reduced growth of liver metastases in preclinical studies. An earlier, smaller phase 2 trial run by the same Leeds team had found that EPA given before surgery had no effect on its primary measure — the rate at which tumour cells were multiplying — yet patients who received it had better overall survival in the 18 months after surgery. EMT2 was designed to test whether that hint of a survival benefit was real, at scale, and over years rather than months.

The survival benefit held up under scrutiny

Within the trial, the overall survival advantage for EPA was not a borderline or fragile finding. It remained consistent in the per-protocol analysis — which looks only at patients who actually took the treatment as planned — where the hazard ratio was 0.60. It also held after statistical adjustment for diabetes, which was twice as common at baseline in the EPA group as in the placebo group, affecting 16 per cent of EPA patients against 8 per cent of those on placebo. Sensitivity analyses pointed in the same direction.

“The overall survival benefit seen with EPA in EMT2, if confirmed in other studies, is clinically meaningful for the group of patients who have had surgery for bowel cancer spread to the liver and in whom the cancer frequently comes back,” Professor Hull said in the ESMO news release. “This is particularly the case as EPA is well-tolerated and safe, even when other cancer treatments are needed.”

Safety was, in fact, one of the cleanest parts of the result. EPA caused no excess of adverse events compared with placebo, no increase in bleeding, and no increase in atrial fibrillation — a heart-rhythm problem that has been a concern with some omega-3 products at high doses. It had no negative impact on quality of life. That matters because any treatment given for years after major surgery must be tolerable, and because many of these patients go on to receive further chemotherapy and other therapies.

One practical wrinkle is that relatively few patients stayed on treatment for the full planned course. Only about 40 per cent of participants in either arm completed treatment, and 22 per cent withdrew from the trial entirely. But withdrawals were balanced between the two groups and were often related to a sense that treatment was futile once the cancer had progressed, or to end-of-life care, rather than to side effects or intolerance, according to Professor Hull. Because the survival analysis counted every randomised patient in the group they were assigned to, regardless of whether they finished treatment, the result is not an artefact of the healthiest patients sticking with the capsules.

The puzzle: longer life, same progression

So why would patients live longer if their cancer returned just as quickly? That is the question every expert at ESMO came back to, and nobody yet has a proven answer.

Professor Hull proposed several possible explanations, all of them biological rather than statistical. One is cachexia — the wasting syndrome of weight loss, loss of appetite and fatigue that afflicts many people with advanced cancer and that itself shortens life. EPA has been studied for years as a possible treatment for cachexia, and if it blunted that wasting, patients might survive longer even while their tumours behaved the same way.

A second possibility is that EPA improved the body’s immune response to the cancer, perhaps through its ability to inhibit cyclooxygenase, an enzyme involved in inflammation. A third is that EPA changed the gut microbiome — the community of bacteria in the bowel — in ways that helped patients tolerate later treatments or resist infection. There is also precedent for the pattern: some cancer immunotherapy trials have shown improved overall survival without a significant impact on progression-free survival, raising the possibility that EPA works, at least in part, through the immune system rather than by directly attacking tumour cells.

Dr Chiara Cremolini, an associate professor of medical oncology at the University of Pisa in Italy who was not involved in the study, called the results “definitely intriguing, from both a biological and a clinical point of view” — and then urged caution. “The lack of concordance between progression-free survival and overall survival results requires careful interpretation,” she said in the ESMO news release. She noted that progression-free survival is a less established indicator of overall survival after surgery for liver metastases than it is after surgery for early-stage bowel cancer, and that additional data would help, including details of the treatments patients received after their cancer returned and the pattern of spread when it did.

Dr Jeffrey Meyerhardt of the Dana-Farber Cancer Institute in Boston, who discussed the trial at ESMO, was blunter about the practical implications. The trial should not change “what we will recommend to patients in clinic next week,” he said. He pointed to a small randomised trial of omega-3 supplements given around the time of colorectal cancer surgery that found numerically worse long-term survival in the supplement arm than in the control arm — a reminder that the evidence base for omega-3 in this setting is not uniformly positive.

Dr Meyerhardt also listed factors that were not fully evaluated in EMT2 and could have influenced the result: how much dietary fibre patients ate, whether they took cyclooxygenase inhibitors such as aspirin or other anti-inflammatory drugs, and possible interactions between EPA and chemotherapy. Another open question, he said, is whether the survival advantage was related to cardiovascular benefits. Omega-3 supplementation, particularly EPA alone, has been shown to reduce cardiovascular mortality, and patients who survive cancer surgery often die of heart disease rather than cancer. If EPA prevented some heart attacks and strokes, that alone could produce a survival gap without any effect on the tumour.

A long road from fish oil to the operating theatre

Omega-3 fatty acids have been one of the most studied — and most argued-over — substances in nutrition and medicine. Found naturally in oily fish such as salmon, mackerel and sardines, and sold widely as dietary supplements, they have been investigated for heart disease, arthritis, depression, dementia and cancer, with results that have often been mixed or modest.

The cardiovascular story is the most developed. The REDUCE-IT trial, published in 2019, showed that 4 grams a day of icosapent ethyl reduced major cardiovascular events in people with high triglycerides and established cardiovascular disease or diabetes. That trial used the same prescription formulation and dose as EMT2. It is also the source of the safety reassurance in EMT2: high-dose prescription EPA is a known quantity in cardiology, with a well-characterised side-effect profile.

But the step from cardiovascular prevention to cancer treatment is a large one, and the history of omega-3 in cancer is littered with promising laboratory findings that did not survive contact with randomised trials. Observational studies have repeatedly linked higher fish consumption or higher blood levels of omega-3 with lower colorectal cancer risk or better outcomes — but people who eat more fish also tend to exercise more, smoke less and eat better overall, and those differences are hard to separate from the fish itself. EMT2 is valuable precisely because it is randomised: the only systematic difference between its two groups was the capsules they took.

That is also why the trial’s funding and disclosures matter. EMT2 was funded by Yorkshire Cancer Research, a charity, and run by an academic team. Amarin provided the EPA and placebo capsules. Professor Hull disclosed financial relationships with the companies SLA Pharma and Curileum. Dr Cremolini disclosed relationships with several pharmaceutical companies, including Amgen, Bayer, Merck and Roche. Dr Meyerhardt reported no relevant disclosures. None of that undermines the result — industry supply of a study drug is routine in academic trials — but it is part of the record, and it is why independent replication by a team with no commercial ties to EPA would carry particular weight.

What this does not mean

It is important to be precise about what EMT2 does and does not show, because the gap between a conference presentation and a change in medical practice is where misunderstanding flourishes.

First, the trial did not show that omega-3 supplements prevent bowel cancer, or that eating more fish will treat it. It tested one specific, highly purified prescription formulation, at a high dose of 4 grams a day, in one specific group of patients — people who had just had liver metastases surgically removed — over two to four years. Ordinary fish-oil capsules sold over the counter contain a mixture of EPA and DHA in widely varying amounts and purity, and they are not the same product that was tested.

Second, the trial missed its primary endpoint. In the hierarchy of clinical trial evidence, that matters. Overall survival was a pre-specified and statistically powered key secondary endpoint — the trial was designed to measure it properly, and the investigators did not go fishing for a positive result after the fact — but a secondary endpoint that succeeds while the primary endpoint fails is always treated with more scepticism than a trial that hits its main target. That is why the experts’ language is so careful: “intriguing”, “if confirmed”, “confirmatory results are likely needed”.

Third, the results have so far been presented at a medical meeting, not published in a peer-reviewed journal. Findings presented at medical meetings should be considered preliminary until they have been through peer review, when independent experts examine the full data, the methods and the analysis. The trial’s rationale and design were published in BMJ Open in 2023, and the full results will need the same scrutiny.

None of that diminishes what was presented in Madrid. A 17-percentage-point gap in five-year survival — 62 per cent against 45 per cent — in a randomised trial of 418 patients is not a small signal, and it appeared in a group of patients for whom the cancer so often returns that any safe, cheap addition to treatment would be welcome. But it is a signal that needs confirming, and a puzzle that needs explaining, before it becomes a prescription.

Analysis: Why It Matters

A cheap, safe drug in a disease with few options

If the EMT2 result is real, its most remarkable feature is how undramatic the intervention is. EPA is not a billion-dollar antibody or a bespoke cellular therapy. It is a purified fatty acid, taken as capsules, with a safety profile already established in cardiology, at a cost far below that of modern cancer drugs. Liver resection for colorectal metastases is a major operation with a high chance of recurrence, and the treatments available after surgery — chemotherapy, further surgery, ablation — are gruelling and expensive. A well-tolerated capsule that added years of life in that setting would be one of the most cost-effective advances in cancer care in years. That is exactly why the result deserves serious, urgent follow-up rather than either hype or dismissal. The history of oncology is full of cheap interventions that were ignored because they were not profitable, and equally full of cheap interventions that turned out not to work. The only way to tell the difference is a confirmatory trial.

The endpoint mismatch is a test of how oncology judges success

EMT2 also exposes a fault line in how cancer trials are judged. Progression-free survival is the workhorse endpoint of modern oncology: it is faster to measure than overall survival, and regulators accept it as a basis for approval. But it is a surrogate — a stand-in for what patients actually care about, which is living longer and living well. EMT2 is a case in which the surrogate and the real outcome pointed in opposite directions. If the overall survival benefit is genuine, then a trial judged solely on its primary endpoint would have been written off as a failure, and a treatment that helps people live longer would have been discarded. That should prompt a serious conversation about whether progression-free survival is the right primary endpoint for trials in the post-surgery setting, where recurrence is common but not immediately fatal, and where patients may live for years with recurrent disease. Dr Cremolini’s observation — that progression-free survival is a less reliable guide to overall survival after liver-metastasis surgery than after early-stage bowel surgery — points in exactly that direction.

Mechanism matters more than usual here

Most of the time, when a cancer drug extends survival, the mechanism is obvious: it shrinks tumours or delays their growth. EMT2 is unusual because the survival benefit arrived without any measurable effect on tumour growth. That makes understanding the mechanism not an academic luxury but a practical necessity. If EPA works by preventing cachexia, it should be tested in other wasting-prone cancers. If it works through the immune system, it might combine with immunotherapy. If it works by preventing cardiovascular deaths, its benefit would be greatest in older patients with heart risk factors, and it would say little about cancer biology at all. Each hypothesis points to a different confirmatory trial, a different patient population and a different way of measuring success. Until the mechanism is pinned down, clinicians cannot know who should receive EPA, when it should start, or how long it should continue — and the trial’s own data cannot answer those questions, because it was not designed to.

Patients will ask about this on Monday — and doctors need a clear answer

The practical consequence of a high-profile conference presentation is immediate: patients who have had liver surgery for bowel cancer, and their families, will read about EMT2 and ask their oncologist whether they should take omega-3. The honest answer, on today’s evidence, is that nobody knows yet — and that taking high-dose supplements without medical advice carries its own risks, including bleeding and interactions with other drugs. That is an unsatisfying answer, but it is the right one, and it places a responsibility on the medical community to move quickly. The worst outcome would be a long period in which patients self-medicate with unstandardised supplements while the confirmatory trial that could settle the question is never funded. The best outcome would be a rapid, publicly funded replication trial, with the mechanism studies built in from the start. The difference between those two futures is a matter of will and money, not science.

A reminder that nutrition is not separate from oncology

Finally, EMT2 is a reminder that the boundary between nutrition and cancer treatment is more porous than either field sometimes admits. Dr Meyerhardt’s closing point at ESMO — that the bar is moving, and that researchers need to keep exploring strategies that are biologically rational, whether they are newer drugs, dietary or lifestyle factors — captures something important. Cancer care has been transformed by targeted drugs and immunotherapy, but the body’s response to cancer is shaped by diet, metabolism, the microbiome and inflammation, all of which are modifiable. A trial that tests a dietary molecule with the same rigour as a new drug is a model for how that territory should be explored: randomised, blinded, placebo-controlled, with hard endpoints and honest reporting of failure as well as success. Whatever EMT2’s final verdict, that approach deserves to become more common, not less.

What to Watch Next

1. Peer-reviewed publication: The full EMT2 results need to appear in a peer-reviewed journal, with complete data on post-progression treatments, patterns of recurrence and causes of death. That publication will allow independent experts to judge whether the survival benefit survives detailed scrutiny.
2. A confirmatory trial: The decisive next step is a second, larger randomised trial testing high-dose EPA after liver resection for colorectal metastases, ideally with the mechanism questions — cachexia, immunity, cardiovascular events — measured from the outset.
3. Mechanism studies: Blood, tissue and microbiome samples collected during EMT2 may reveal whether EPA changed immune function, inflammation or body composition. Any such finding would help explain the survival gap and identify the patients most likely to benefit.
4. Guideline responses: Watch for reactions from bodies such as the National Institute for Health and Care Excellence, the European Society for Medical Oncology and surgical oncology societies. None is likely to recommend EPA on the basis of a single conference presentation, but their assessments will shape whether a confirmatory trial is funded.
5. The supplement market: Expect a surge of marketing claiming that fish-oil capsules fight cancer. Patients should note that EMT2 tested a prescription-grade purified EPA at 4 grams a day under medical supervision — not over-the-counter supplements — and that no health authority has endorsed omega-3 as a cancer treatment.

Sources

• ESMO Congress 2026 news release: “Phase 3 EMT2 trial demonstrates survival benefit for high-dose EPA” (News-Medical, 26 October 2026)
• Medscape: “Can Omega-3 Supplements Prolong Life in Metastatic CRC?” (28 October 2026)
• Drugs.com / HealthDay: “Omega-3s Might Improve Survival In Advanced Cancer, Trial Finds” (30 October 2026)
• BMJ Open: EMT2 trial rationale and design (2023)
• ClinicalTrials.gov: EMT2 trial registry entry (NCT03428477)