On Oct. 8, the US Food and Drug Administration approved an immunotherapy for people whose colon cancer has already been surgically removed but may come back. Atezolizumab, sold by Genentech as Tecentriq, can now be added to standard chemotherapy for stage III colon cancer when the tumor is mismatch repair deficient, known as dMMR, according to AJMC's report on the approval. The trial behind it shows a large drop in recurrence. It does not yet show that patients live longer.

That distinction matters, and it is easy to miss in the headline.

What the FDA Approved

The approval covers atezolizumab with a fluoropyrimidine and oxaliplatin as adjuvant treatment, meaning treatment after surgery, for stage III dMMR colon cancer in adults and in pediatric patients aged 2 and older. The agency also approved a subcutaneous version, atezolizumab and hyaluronidase (Tecentriq Hybreza), given as a single injection every three weeks for patients aged 12 and older who weigh at least 40 kilograms.

dMMR means the tumor cells cannot fix errors that arise when DNA is copied. These tumors accumulate mutations, which tends to make them visible to the immune system, and they respond to drugs that release the brakes on immune cells. Atezolizumab is one of those drugs.

What the ATOMIC Trial Showed

The approval rests on ATOMIC, a randomized phase 3 trial of 711 adults and one pediatric patient with stage III dMMR colon cancer who had undergone complete tumor removal. Half received atezolizumab plus a chemotherapy regimen called mFOLFOX6 for 12 cycles, followed by six months of atezolizumab alone. The other half received mFOLFOX6 alone.

The main result was disease-free survival. The hazard ratio was 0.50 (95% confidence interval 0.35 to 0.73), meaning half the hazard of recurrence or death compared with chemotherapy alone. In the published paper in the New England Journal of Medicine, led by Frank Sinicrope, 3-year disease-free survival was 86.3% with atezolizumab and 76.2% without, after a median follow-up of 40.9 months.

In plain terms, about 10 more patients per 100 were free of cancer at three years.

The Survival Question

Overall survival did not differ significantly. The hazard ratio was 0.90 (95% confidence interval 0.55 to 1.47, P = 0.68), with 5-year survival rates of 89.7% and 87.9%, according to the same report. The confidence interval is wide, so the trial cannot rule out a survival benefit or a small harm.

Several explanations are plausible. Most patients with stage III disease survive five years with surgery and chemotherapy alone, so deaths are few and differences take time to show. Follow-up is still relatively short. Patients whose cancer returned in the chemotherapy-only group may also have received immunotherapy afterward, which would narrow any gap in survival. Those are my observations on how trials like this usually behave, not statements from the paper.

My view: the approval is reasonable, since delaying or preventing recurrence matters to patients, but it rests on a surrogate endpoint. The survival data are the number to watch as follow-up grows.

The Cost in Side Effects

Adding immunotherapy increased toxicity. Grade 3 or higher adverse events occurred in 84.1% of patients on the combination and 71.9% on chemotherapy alone, according to the trial report. Two treatment-related deaths occurred in the combination arm.

The prescribing information carries warnings for immune-mediated reactions, in which the immune system attacks healthy organs, as well as infusion reactions, complications after allogeneic stem cell transplantation, and embryo-fetal toxicity. Patients taking checkpoint inhibitors need monitoring for these effects.

One Pediatric Patient and a Global Review

Two details stand out. First, the trial enrolled just one pediatric patient, yet the label covers patients as young as 2 for the intravenous drug. The sources I reviewed do not explain the agency's reasoning, which may rest on evidence from other cancers and extrapolation.

Second, the FDA reviewed the applications under Project Orbis, an initiative that allows concurrent review of cancer drugs with partner regulators. Australia's Therapeutic Goods Administration, Health Canada, Israel's Ministry of Health and Switzerland's Swissmedic took part, and their reviews are continuing.

What It Means for Patients

For patients and doctors the immediate step is testing. Deciding who qualifies depends on knowing whether the tumor is dMMR, which is determined by laboratory testing of the surgical sample. Only a minority of colon cancers have this feature, so most patients with stage III disease will not be eligible.

The discussion echoes questions Blogerroom has covered elsewhere about matching cancer treatment to biology, such as the mesothelioma drug that turns a tumor's defenses against it and the urine RNA test that aims to predict response to bladder cancer immunotherapy. In each case the promise is the same, and so is the caution: a strong biomarker result still has to prove itself in survival.

Anyone with a colon cancer diagnosis should ask their oncologist whether tumor testing for mismatch repair status has been done, and how the benefits and side effects apply to their case. This article is general information, not medical advice.