Omega-3 polyunsaturated fatty acid eicosapentaenoic acid (EPA) versus placebo after resection of colorectal cancer liver metastases (EMT2): A multicentre, double-blind, randomised phase 3 trial

EMT2

Patient / Population Intervention / Exposure Comparison Outcome

In patients undergoing resection of colorectal cancer liver metastases, eicosapentaenoic acid did not improve progression-free survival compared with placebo, although overall survival, a secondary endpoint, was nominally improved.

N

418patients

Design

Multicentre, double-blind, randomised phase 3 trial that allocated adults undergoing resection of colorectal cancer liver metastases with curative intent 1:1 by minimisation to pure EPA ethyl ester or mineral oil placebo capsules, started at least one day before surgery and continued for two to four years (EudraCT 2016-000628-24; NCT03428477).

Endpoint

Progression-free survival.

ResultMedian PFS 13.5 (95% CI 11.8-17.9) months with EPA vs 12.9 (95% CI 10.4-16.3) months with placebo (HR 0.89 [95% CI 0.70-1.13]; P=0.33). Five-year OS 62.0% (95% CI 52.7-70.0) with EPA vs 45.1% (95% CI 33.7-55.9) with placebo (HR 0.68 [95% CI 0.48-0.96]; P=0.030). In the safety population (n=411) there was no excess of total adverse events, bleeding or atrial fibrillation with EPA. Quality of life measures were similar between groups.

Hull MA, Ow PL, Smith AF, et al. Presented at ESMO Congress 2026, abstract LBA30, 25 October. Not yet published.

Discussion & critique

This double-blind, placebo-controlled trial shows that EPA did not improve its primary endpoint, progression-free survival. The overall survival gain is a key secondary endpoint, and the abstract describes no hierarchical testing, so its P value is nominal and the finding is hypothesis-generating. The abstract does not report median follow-up, the number of deaths, or treatment after progression, and it does not explain how survival could improve without any effect on progression. The confidence intervals around five-year survival are wide, and safety is summarised only in qualitative terms.