NICHE: What the Colorectal Cancer Trial Found

The question NICHE asked

Immune checkpoint drugs work very well in advanced colorectal cancer with one particular feature. Without it, they hardly work at all. NICHE asked a follow-on question: would giving those drugs before surgery, in early-stage colon cancer, change that picture?

The trial was registered as NCT03026140 and run at the Netherlands Cancer Institute. Its results were published in Nature Medicine in April 2020. It was explicitly exploratory, and the researchers said so in the paper.

Mismatch repair, in plain terms

Cells copy their DNA constantly, and copying makes mistakes. A set of proteins called the mismatch repair system finds and fixes them.

When that system is broken, errors pile up. Tumors with a broken system are called mismatch repair-deficient, or dMMR. Tumors with a working system are mismatch repair-proficient, or pMMR.

The difference matters for immunotherapy. A dMMR tumor accumulates so many mutations that its cells start producing abnormal proteins the immune system can recognize as foreign. A pMMR tumor looks far more like normal tissue, and the immune system largely ignores it.

Whether a tumor is dMMR or pMMR is determined by laboratory testing. Our page on biomarker testing covers how those results are produced and what they are used for.

What participants received

Forty patients with colon cancer that had not spread were treated. Twenty-one had dMMR tumors and 20 had pMMR tumors. They received a single dose of ipilimumab and two doses of nivolumab before their operation. The pMMR group was split again by whether they also got celecoxib. Three patients in an initial safety run-in received nivolumab alone.

Ipilimumab and nivolumab release two different brakes on the immune system, CTLA-4 and PD-1. Our overview of immunotherapy explains how that class of drug works.

Of those who received both drugs, 20 with dMMR tumors and 15 with pMMR tumors, 35 could be assessed for effect.

The primary endpoint was not what the headlines covered

The trial's stated primary objective was safety and feasibility. In a study giving new drugs shortly before an operation, the first thing to establish is that the operation still happens on time.

It did. Treatment was well tolerated, and every patient underwent radical surgery without delay. That is the endpoint the trial was designed around, and it was met.

What the tumors did

The efficacy results were the reason NICHE became widely known.

In dMMR tumors, pathological response occurred in 20 of 20. That is 100 percent, with a 95 percent exact confidence interval of 86 to 100 percent. Nineteen were major responses, meaning 10 percent or less living tumor remained. Twelve had no living tumor left at all.

In pMMR tumors the picture was entirely different: 4 of 15, or 27 percent, with a 95 percent exact confidence interval of 8 to 55 percent. Three were major responses and one was partial.

The researchers also found that the presence of a particular immune cell, the CD8+PD-1+ T cell, predicted which pMMR tumors would respond.

What a "pathological response" is, and is not

After surgery, the removed tissue goes to a pathologist, who measures how much living tumor is left. Less living tumor means the drugs did something. That is a real, physical observation, not a shadow on a scan.

But it is a surrogate endpoint. It stands in for what people actually want to know: whether they live longer without the cancer returning. NICHE had no survival data at publication. The authors said so directly. This approach could become standard care for a defined group, they wrote, only when validated in larger studies with at least three years of disease-free survival data.

That sentence is the honest core of the paper, and it is worth more than the 100 percent figure.

Where colon cancer sits statistically

The American Cancer Society projects 158,850 new colorectal cancers in the United States in 2026 and 55,230 deaths; those appear on SEER, the federal cancer statistics program. It is the fourth most common cancer in the country. SEER's own staging data show about 34 percent are found while still confined to the bowel. That is roughly the group NICHE studied.

Five-year relative survival across all stages was 65.4 percent for people diagnosed from 2016 through 2022. That is a group statistic covering every stage and every treatment era, and it forecasts nothing for an individual.

Only a minority of colon cancers are dMMR, which means the striking result applies to a small slice of an already specific group.

What this trial cannot tell you

• It was small and exploratory. Thirty-five patients were evaluable, and the confidence interval for the pMMR group runs from 8 percent to 55 percent, which is very wide.
• There was no randomization and no comparison group. Nobody in NICHE received standard care alone for comparison.
• Pathological response is not survival. No survival data existed at publication.
• Trial participants meet specific entry criteria, so results may not carry over to everyone with the same diagnosis.
• NICHE did not change practice on its own. It set up the larger NICHE-2 trial, and the registry record shows the NICHE program still enrolling.

Our pages on clinical trial phases and what standard of care means in a trial explain why an exploratory study and a practice-changing one are read so differently.

Sources

• https://pubmed.ncbi.nlm.nih.gov/32251400/
• https://clinicaltrials.gov/api/v2/studies/NCT03026140
• https://www.cancer.gov/types/colorectal
• https://seer.cancer.gov/statfacts/html/colorect.html

An AI-assisted editorial system helped prepare this page. No named medical reviewer has reviewed it unless one is listed.

See an error, old source, or unclear wording? Tell us.