NICHE-2 dMMR Colon Cancer: Immunotherapy Results for USMLE

A long-term analysis of the phase 2 NICHE-2 trial, published in The Lancet, reports no disease recurrences among 111 patients with locally advanced, nonmetastatic mismatch repair–deficient (dMMR) colon cancer. The patients received only about four weeks of neoadjuvant nivolumab plus ipilimumab before surgery. All were alive at last follow-up, with a median postoperative follow-up of 58.7 months. The result is striking, but it comes from a single-arm trial, so it is not proof of cure.

What happened

NICHE-2 was a single-arm, multicenter, phase 2 study. It tested neoadjuvant nivolumab (anti-PD-1) plus ipilimumab (anti-CTLA-4) in patients with nonmetastatic dMMR colon cancer. Patients received two cycles of immunotherapy over roughly four weeks and then underwent surgery.

Earlier reports from the trial showed a pathological response in 98% of patients. A pathological complete response, meaning no viable tumor cells in the resected specimen, occurred in 68%. The three-year analysis had already shown no recurrences.

The new analysis extends follow-up to a median of 58.7 months. Among the 111 patients in the primary efficacy population:

• No recurrences were observed.
• All patients were alive at their most recent follow-up.
• Three-year disease-free survival was 100%.
• Estimated five-year overall survival was also 100%.

The investigators also studied circulating tumor DNA (ctDNA). Tumor-derived DNA was detectable in the blood of most patients before treatment. Among patients with detectable baseline ctDNA and available follow-up samples, ctDNA cleared in 45% after the first cycle and in 82% after the second, before surgery. However, preoperative ctDNA testing was not sensitive enough to reliably identify every patient with a pathological complete response.

Why it matters

Many patients in NICHE-2 had locally advanced tumors with a higher risk of recurrence. Seeing no relapses after a brief, roughly four-week treatment is unusual for this setting. According to the report, NICHE-2 results have already contributed to international guidelines. Several countries, including Norway, Scotland and Uruguay, have adopted the approach, and efforts are under way to make it available in the Netherlands.

The study also raises a practical question: could surgery eventually be avoided in selected patients who have a complete response? At present, a complete response can only be confirmed on the surgical specimen. The researchers are exploring how ctDNA, endoscopy and imaging might be combined to assess response without surgery. Other studies are testing whether a single immunotherapy agent could work as well and whether longer treatment would raise the complete response rate. These remain open research questions, not established practice.

What to know for the USMLE

This news ties together several high-yield topics: DNA repair, hereditary cancer syndromes, tumor immunology and checkpoint pharmacology.

• Mismatch repair (MMR) fixes base-pair mismatches and small insertion-deletion loops that arise during DNA replication. The key proteins are MLH1, MSH2, MSH6 and PMS2. MSH2 pairs with MSH6, and MLH1 pairs with PMS2.
• dMMR leads to microsatellite instability (MSI-H). Without repair, errors accumulate in repetitive microsatellite sequences. Tumors are tested by immunohistochemistry for MMR proteins or by PCR-based MSI testing.
• Lynch syndrome (hereditary nonpolyposis colorectal cancer, HNPCC) is an autosomal dominant germline mutation in an MMR gene. It increases the risk of colorectal cancer, usually proximal, as well as endometrial cancer. Ovarian, gastric, urinary tract and other cancers are also associated. Think of it when a vignette describes early-onset colon cancer with a strong family history (Amsterdam criteria).
• Sporadic dMMR tumors usually result from MLH1 promoter hypermethylation, often alongside a BRAF V600E mutation. Lynch-associated tumors typically lack BRAF V600E and MLH1 hypermethylation, so this distinction is a classic exam point.
• Why immunotherapy works: dMMR tumors carry a high mutational burden, which produces many neoantigens and abundant tumor-infiltrating lymphocytes. Tumors suppress T cells through PD-L1 binding PD-1 on T cells. Checkpoint blockade releases that brake.
• Drug mechanisms: Nivolumab and pembrolizumab block PD-1. Ipilimumab blocks CTLA-4, which acts earlier, mainly during T-cell priming. Pembrolizumab has a tissue-agnostic indication for MSI-H/dMMR solid tumors.
• Immune-related adverse events follow from generalized T-cell activation. They include colitis, pneumonitis, hepatitis, dermatitis, and endocrinopathies such as thyroiditis and hypophysitis. Combination anti-PD-1/anti-CTLA-4 therapy is generally associated with more of these than monotherapy. The source material here does not report safety data, so don't cite figures from this article.
• Pathology clues: dMMR colon cancers are often right-sided and mucinous or poorly differentiated, with a prominent lymphocytic infiltrate (such as a Crohn-like reaction). They typically carry a better stage-adjusted prognosis than microsatellite-stable tumors. A related classic point is that stage II dMMR tumors derive little benefit from 5-fluorouracil alone.
• ctDNA is tumor-derived DNA in plasma. It is being studied as a marker of response and minimal residual disease, but the study found preoperative testing is not yet sensitive enough to replace pathology.

Limitations

• Single-arm, phase 2 design. There was no randomized comparison group, so the trial cannot show how outcomes compare with surgery alone or with other neoadjuvant strategies.
• Primary efficacy population. The 100% figures apply to the 111 patients in this defined population, in a trial restricted to nonmetastatic dMMR colon cancer. They do not apply to microsatellite-stable colon cancer.
• Follow-up is long but finite. A median of 58.7 months means some patients were followed for less time. The five-year overall survival figure is described as an estimate. Late recurrences cannot be excluded with certainty, and "cure" is too strong a word.
• Surgery was still part of treatment. The 68% pathological complete response was confirmed only after resection. Whether surgery can safely be skipped is unanswered.
• ctDNA is not ready to guide decisions. Preoperative ctDNA testing could not reliably identify all complete responders.
• Questions about the regimen remain open. The trials of monotherapy and longer treatment are still ongoing.

This article is educational and is not guidance for treating any individual patient.

FAQ

What is dMMR colon cancer?

It is colon cancer whose cells cannot properly repair DNA replication errors because of a defective mismatch repair system. This causes microsatellite instability (MSI-H) and a high mutation burden. It can be sporadic or caused by Lynch syndrome.

Why do dMMR tumors respond to anti-PD-1 therapy?

Their many mutations generate abundant neoantigens, so the immune system recognizes them as foreign. Blocking PD-1 (and CTLA-4) releases the checkpoint restraint on T cells, which can then attack the tumor.

Does NICHE-2 mean surgery is no longer needed for dMMR colon cancer?

No. All patients in NICHE-2 had surgery, and a complete response could be confirmed only on the surgical specimen. Researchers are investigating whether ctDNA, endoscopy and imaging could someday identify patients who may safely avoid surgery.

Key takeaways

• In the single-arm phase 2 NICHE-2 trial, none of 111 patients with nonmetastatic dMMR colon cancer had a recurrence after about four weeks of nivolumab plus ipilimumab and surgery (median follow-up 58.7 months), and all were alive at last follow-up.
• For exams, link dMMR/MSI-H to MLH1, MSH2, MSH6, PMS2, Lynch syndrome, MLH1 hypermethylation/BRAF V600E in sporadic cases, and anti-PD-1/CTLA-4 therapy.
• The data are impressive but come without a control group. Omitting surgery and using ctDNA to guide decisions are still research questions.