A new immune partner for stage III colon cancer

The hospital corridor after colon surgery can hold two kinds of waiting: recovery from the operation, and the uneasy months before the next scan. The phase 3 ATOMIC trial tested whether adding checkpoint blockade to chemotherapy could reduce that second uncertainty for one molecularly defined group. Evidence checked 3 November 2026.

At a glance

The phase 3 ATOMIC trial compared adjuvant modified FOLFOX6 alone with the same chemotherapy plus atezolizumab in people with resected stage III colon cancer whose tumors were deficient in mismatch repair (dMMR). The primary outcome, disease-free survival, favored the combination. The finding is a major advance for a subgroup often less responsive to chemotherapy alone, while overall-survival follow-up and careful assessment of immune-related harms remain important. In the peer-reviewed NEJM report, 3-year disease-free survival was 86.3% versus 76.2% after median follow-up of 40.9 months; grade 3 or 4 adverse events occurred in 84.1% versus 71.9%.

Why does the repair system matter?

Imagine a cell copying a manuscript billions of letters long. DNA mismatch-repair proteins are the proofreaders: they recognize many copying errors and help correct them. When that machinery is missing, errors accumulate. The resulting tumor can carry many mutations and may display abnormal proteins that the immune system can recognize. Yet cancers also learn to hide from immune attack. One common disguise is the PD-L1 signal, which can engage PD-1 on immune cells and dampen their response.

The biology makes an immune checkpoint inhibitor plausible. But plausibility is not evidence, and earlier clinical results have differed by stage and treatment setting. ATOMIC asked a practical question: after surgeons remove visible stage III disease, can immunotherapy added to established chemotherapy reduce the chance that microscopic disease returns?

What did ATOMIC test?

In this phase 3 randomized study, 712 patients with resected stage III dMMR colon cancer were assigned to modified FOLFOX6 for six months, either alone or with atezolizumab. The combination arm continued atezolizumab after chemotherapy, for a total treatment period of about one year. Disease-free survival—time until recurrence, a new cancer, or death—was the primary endpoint. The peer-reviewed paper reports a 3-year disease-free survival estimate of 86.3% with atezolizumab plus chemotherapy versus 76.2% with chemotherapy alone; the hazard ratio was 0.50. These are group-level trial estimates, not a personal forecast.

The randomized comparison and prespecified endpoint strengthen the result. The study was published in The New England Journal of Medicine in March 2026, led by Frank A. Sinicrope and colleagues: ATOMIC trial report. The trial registry gives the protocol and eligibility framework. Atezolizumab blocks PD-L1; it is not a chemotherapy drug, and the combination brings distinct toxicities as well as potential benefit.

What does “disease-free” mean for a patient?

It means the trial counted recurrences and deaths over time, not that participants were guaranteed never to develop cancer again. A hazard ratio compares the rate at which events occurred across follow-up; it is not the same thing as saying every patient’s risk was cut exactly in half. Nor does an early disease-free-survival advantage automatically prove an overall-survival advantage. Longer follow-up can show whether fewer recurrences translate into people living longer, and how often treatment causes lasting immune effects.

Adjuvant treatment also has a special emotional burden. After surgery, scans may show no visible cancer, yet doctors recommend months of treatment to reduce an uncertain future risk. A trial result can improve that decision without eliminating uncertainty. Patients and clinicians still need to weigh tumor biology, stage, health, side effects, and local standards of care.

Why could it matter beyond one study?

The result adds evidence that immunotherapy can help in a curative-intent setting for a molecularly selected colon cancer population. It may sharpen future questions: can some patients safely receive less chemotherapy? Which biomarkers identify the greatest benefit? What is the best duration of checkpoint blockade? Those are new trials, not conclusions already delivered by ATOMIC.

For African health systems, the pathway from result to routine care includes more than drug availability. Reliable pathology to identify dMMR, oncology teams to monitor immune toxicities, infusion capacity, and sustainable financing all matter. The biology is relevant wherever colon cancer occurs, but the trial does not establish affordability, access, or outcomes in African populations. Local researchers and health ministries can use it as a prompt to strengthen molecular diagnosis and build evidence on implementation.

What are the limits?

ATOMIC enrolled a specific population: patients with stage III colon cancer after surgery and dMMR tumors. It should not be generalized to every colon cancer, rectal cancer, metastatic disease, or every checkpoint inhibitor. Follow-up is still needed to understand overall survival, late toxicities, and whether benefits vary across patient subgroups. The trial was not designed to answer whether immunotherapy alone can replace chemotherapy.

The best headline is therefore measured: adding atezolizumab to FOLFOX improved disease-free survival in this defined phase 3 population. That is meaningful progress. It is not a universal prescription or proof of cure.

Questions readers ask

Is atezolizumab approved for this use everywhere?

Approval and guideline status depend on jurisdiction and can change. The trial result itself is not a regulatory approval. Check the relevant national regulator and treating oncology team for current local status.

Does dMMR mean the cancer is inherited?

Not necessarily. Mismatch repair can be lost through inherited variants or changes acquired within the tumor. Clinical genetics assessment may be appropriate, but a tumor result alone does not establish a family diagnosis.

Does this mean chemotherapy is no longer needed?

No. ATOMIC tested atezolizumab added to chemotherapy, followed by atezolizumab. It did not test replacing chemotherapy with the immune drug.

Sources

• Sinicrope et al., ATOMIC, NEJM (2026)
• ATOMIC trial record, ClinicalTrials.gov
• ASCO Post summary of the ATOMIC results

Continue exploring

The pancreatic cancer pill that finally corners a slippery target. DESTINY-Lung04 reports a first-line result for HER2-mutant lung cancer.

Cover: Illustrative micrograph of human colon cancer cells. Credit: National Cancer Institute Center for Cancer Research. Public domain. This is not an image from the ATOMIC trial. Credit: National Cancer Institute Center for Cancer Research. Public domain (U.S. federal government work). Original source.