Immunotherapy for stage III colon cancer improved 3-year disease-free survival

Immunotherapy for stage III colon cancer produced a substantial benefit even after the visible tumor had already been surgically removed. In a randomized phase 3 trial limited to tumors with deficient mismatch repair, or dMMR, adding atezolizumab to standard chemotherapy increased the proportion of patients alive and disease-free at three years from 76.2% to 86.3%.

That is an absolute difference of 10.1 percentage points. Put another way, roughly ten additional patients out of every hundred were free of recurrence or death at three years in the atezolizumab group.

The ATOMIC trial enrolled 712 patients after complete resection of stage III colon cancer. The defining feature was molecular: every locally identified tumor had deficient DNA mismatch repair.

That biological defect is central to why immunotherapy was tested in the first place.

Turning genomic instability into a vulnerability

Every time a cell divides, DNA replication introduces errors. The mismatch-repair machinery normally detects and corrects many of them. In dMMR cancers, that repair system is defective.

As a result, tumor cells accumulate unusually large numbers of mutations. Those mutations can generate abnormal proteins that make the cancer more visible to the immune system. At the same time, tumors can exploit immune checkpoints to suppress attacking T cells.

Atezolizumab blocks PD-L1, removing part of that inhibitory signal.

Checkpoint inhibitors are already highly effective in some patients with metastatic dMMR or MSI-high colorectal cancer. ATOMIC addressed a different setting: could immunotherapy help much earlier, after surgery, when there is no visible tumor but microscopic residual cancer may remain and later cause recurrence?

The trial randomly assigned 355 patients to six months of modified FOLFOX6 — fluorouracil, leucovorin and oxaliplatin — plus atezolizumab, followed by another six months of atezolizumab alone.

Another 357 patients received six months of mFOLFOX6 without immunotherapy.

The population was not restricted to favorable disease. Overall, 53.9% of tumors had high-risk T4 or N2 features, or both.

A ten-point separation at three years

After a median follow-up of 40.9 months, there had been 46 disease-free survival events in the atezolizumab group and 81 in the chemotherapy-alone group.

Three-year disease-free survival was 86.3% with atezolizumab plus chemotherapy and 76.2% with chemotherapy alone.

The hazard ratio for recurrence or death was 0.50. That represents roughly a halving of the relative event hazard during follow-up; it should not be confused with the absolute difference between the groups, which was about ten percentage points at three years.

The result was similar when the analysis was restricted to patients whose dMMR status was confirmed centrally.

Overall survival, however, has not yet shown a significant difference. With median follow-up of 45.8 months for that endpoint, there had been 31 deaths in the atezolizumab group and 33 in the control group. Estimated five-year overall survival was 89.7% versus 87.9%, respectively, with a hazard ratio of 0.90.

The trial therefore provides strong evidence for improved disease-free survival, not yet for longer life.

Better disease control came with more toxicity

The atezolizumab regimen was also more intensive and lasted longer. Grade 3 or 4 adverse events occurred in 84.1% of patients receiving atezolizumab plus mFOLFOX6 compared with 71.9% receiving chemotherapy alone.

Grade 3 or 4 fatigue occurred in 10.1% versus 3.3%, and severe reductions in neutrophil counts in 43.6% versus 35.9%. Several suspected immune-related effects — including hypothyroidism, hyperglycemia, colitis and rash — were also more common with atezolizumab.

Six grade 5 adverse events occurred in the atezolizumab arm and two in the chemotherapy arm. Investigators considered two deaths in the atezolizumab group, one from sepsis and one sudden death, to be treatment related.

So the ATOMIC result is not simply a story of adding an immune drug and obtaining a free improvement. The gain in disease control came with greater treatment burden, more severe toxicity and a total immunotherapy duration of one year.

The deeper lesson lies in the tumor test that came before treatment. Surgery removed the visible cancer in every patient. Testing for dMMR then identified a biological weakness in some of those cancers that could still be targeted after the operation.

In this population, a molecular feature of the removed tumor helped determine how aggressively to treat what could no longer be seen — microscopic disease that might otherwise become the next recurrence.