IPM Take
This is one of the most important precision-oncology fights: not finding a new drug, but deciding who actually needs an old one. ctDNA could help spare some patients unnecessary chemotherapy while identifying others at high risk. But if health systems adopt it without clear evidence, reimbursement and counselling frameworks, precision medicine becomes another expensive uncertainty machine.
Executive Summary
Findings from an overall-survival analysis of the Phase II GALAXY study, presented at ESMO GI 2026 and reported during the July window, suggest that postoperative ctDNA may help predict benefit from adjuvant chemotherapy after resection of colorectal liver metastases. The analysis included 298 patients with ctDNA results available 2 to 10 weeks after surgery, assessed using a tumour-informed assay. In the upfront-surgery cohort, molecular residual disease positivity was associated with substantially worse outcomes. Four-year overall survival was 65% among ctDNA-positive patients who received adjuvant chemotherapy versus 33% among those who did not. Experts cautioned that the evidence is promising but not yet strong enough for routine use outside clinical trials.
Why it matters
- Patients / advocates: The right test could reduce unnecessary treatment, but a premature test could create anxiety without clear action.
- Clinicians: ctDNA may sharpen postoperative risk assessment, especially when the decision to give chemotherapy is uncertain.
- Diagnostics / pathology: Molecular residual disease testing is moving from research into pathway design, and quality control will matter.
- Payers / HTA bodies: Coverage decisions need evidence that ctDNA-guided therapy improves outcomes, not only that it predicts risk.
The chemotherapy question after colorectal liver-metastasis surgery has always been uncomfortable.
Some patients need it. Some may not. Almost everyone pays a price if the decision is wrong.
That is why the GALAXY data matter. The updated analysis suggests postoperative ctDNA can identify patients at high risk after resection of colorectal liver metastases and may help predict who benefits from adjuvant chemotherapy. In the upfront-surgery cohort, ctDNA-positive patients who received adjuvant chemotherapy had a four-year overall survival rate of 65%, compared with 33% among those who did not. Disease-free survival was also sharply different, at 38%versus 7%.
This is exactly the kind of evidence personalised medicine was supposed to generate: not more treatment for everyone, but better treatment decisions for specific patients.
Still, this is not a green light for uncontrolled adoption.
The caution from outside experts is important. The evidence is promising, but it is not yet sufficient to make ctDNA-guided adjuvant treatment routine outside clinical trials. Randomised studies are still needed. A positive ctDNA result can identify danger, but danger is not the same as proof that a specific intervention will fix it.
This distinction matters because diagnostics can move into clinical practice faster than policy can keep up.
Who pays for postoperative ctDNA? Which assay is acceptable? How quickly must the result return? What does a clinician tell a patient with a positive result when the evidence is still developing? What happens in systems where ctDNA is available privately but not publicly reimbursed?
Those are not technical questions. They decide whether precision oncology becomes equitable medicine or concierge medicine.
The most powerful promise here is also the most politically sensitive: using molecular residual disease to stop overtreating some patients and intensify treatment for others. That is the future. But it has to arrive through evidence, not enthusiasm alone.
