A short course of immunotherapy given before surgery eliminated detectable colon cancer in a majority of patients whose tumors lacked a key DNA repair mechanism, according to results published in the New England Journal of Medicine. The findings, drawn from the NICHE-2 trial and a separate pembrolizumab study, showed that patients with mismatch repair-deficient (dMMR) colon cancer achieved pathologic complete responses at striking rates, meaning surgeons found no residual tumor when they operated. The results raise a pressing question: whether some of these patients could eventually skip surgery entirely.
Why pre-surgery immunotherapy for dMMR colon cancer demands attention now
Standard treatment for locally advanced colon cancer still sends nearly all patients straight to the operating room, followed by months of chemotherapy. That protocol does not distinguish between tumors that carry a mismatch repair deficiency and those that do not. Yet roughly 10 to 15 percent of colon cancers are dMMR, a molecular profile that makes them highly sensitive to immune checkpoint inhibitors. The disconnect between biology and current practice is what makes these trial results so consequential.
The NICHE trial platform, registered on ClinicalTrials.gov, tested neoadjuvant nivolumab plus ipilimumab in early-stage colon cancer. Its second phase, NICHE-2, focused on dMMR tumors and produced very high pathologic response rates, with many patients showing complete tumor elimination at the time of surgery. A separate single-arm trial tested just one cycle of pembrolizumab before surgery in patients with stages I through III dMMR colon cancer and also reported pathologic complete responses.
These results did not emerge in a vacuum. An earlier study of PD-1 blockade in mismatch repair-deficient rectal cancer had already demonstrated that immunotherapy could produce complete clinical responses in a related tumor type. That rectal cancer work, published in the New England Journal of Medicine under DOI 10.1056/NEJMoa2201445, established the biological plausibility that checkpoint inhibitors could erase dMMR gastrointestinal tumors before a scalpel ever touched them. The colon cancer findings now extend that principle to a larger patient population with a different anatomical site.
A hypothesis gaining traction among researchers is whether circulating tumor DNA (ctDNA) clearance after just one cycle of PD-1 blockade could predict which patients will achieve a complete pathologic response. If validated, ctDNA testing could function as a real-time triage tool, identifying dMMR patients whose tumors have already been destroyed by immunotherapy and who might safely forgo surgery altogether. That would represent a fundamental shift in how colon cancer is managed, moving from a one-size-fits-all surgical approach to a biomarker-guided strategy.
Trial results that redraw the treatment map for dMMR tumors
The New England Journal of Medicine reported that neoadjuvant immunotherapy in locally advanced dMMR colon cancer led to very high response rates, including pathologic complete response in a large share of participants, in a study of preoperative checkpoint blockade. Surgery generally proceeded without complications in patients who received the pre-operative immunotherapy, and the absence of residual cancer in surgical specimens was confirmed by standard pathological review.
The pembrolizumab trial, registered as NCT05662527, took a minimalist approach: a single cycle of the PD-1 inhibitor before surgery. Even with that brief exposure, the trial reported pathologic complete responses in patients with stages I–III dMMR colon tumors. The fact that one dose could trigger such deep responses speaks to how vulnerable dMMR tumors are to immune attack once the PD-1 brake is released.
Long-term follow-up data from the NICHE-2 cohort, reported in Nature, added another dimension. That analysis tracked dMMR patients who received neoadjuvant nivolumab plus ipilimumab and found durable outcomes over extended observation, with low rates of recurrence emerging during the available follow-up period. The same Nature publication also explored whether the approach worked in mismatch repair-proficient (pMMR) colon cancers and found that the same dramatic benefit did not appear in that group. This biological divide is central to the story: the therapy works because dMMR tumors carry a high burden of mutations that the immune system can recognize, while pMMR tumors lack that vulnerability.
The convergence of these independent datasets, from different drugs, different dosing schedules, and different institutions, strengthens the case that the effect is real and reproducible. Two checkpoint inhibitors in combination and a single agent alone both produced the same pattern of deep responses in the same molecular subtype. For clinicians, that consistency makes it harder to dismiss the results as a fluke or a quirk of trial design.
Gaps in the evidence that patients and doctors still face
For all the excitement, several critical questions remain unanswered. The published trials report aggregate pathologic response rates but do not yet provide granular patient-level data broken down by tumor stage, anatomical location within the colon, or patient comorbidities. A 70-year-old patient with a right-sided stage III tumor and diabetes may respond differently than a 45-year-old with a stage I left-sided lesion, yet the current evidence base does not fully disentangle those scenarios.
Another limitation is the relatively short follow-up for many patients treated in the neoadjuvant setting. Pathologic complete response is a powerful surrogate marker, but it is not a guarantee of cure. Late recurrences after apparent eradication of disease are well documented in other cancers, and only years of observation will show whether the early success of pre-surgery immunotherapy in dMMR colon cancer translates into sustained freedom from relapse and improved overall survival.
There is also no consensus yet on how to integrate ctDNA into routine decision-making. While small datasets suggest that ctDNA clearance after immunotherapy may track with deep tumor regression, assays differ in sensitivity, and the optimal timing of blood draws has not been standardized. Without harmonized testing methods and validated thresholds, ctDNA risks becoming a promising but unreliable compass for steering patients away from surgery.
Equally unsettled is what to do for dMMR patients who do not achieve a complete response. The trials to date have focused on those who respond dramatically, but a subset of tumors either shrink only partially or fail to respond at all. For these patients, it is unclear whether neoadjuvant immunotherapy adds meaningful benefit or simply delays definitive surgery and chemotherapy. Understanding predictors of resistance within the dMMR population will be essential before this strategy can be applied broadly.
What this could mean for the future standard of care
Despite these uncertainties, the implications of the current data are hard to ignore. If further studies confirm that a short course of immunotherapy can reliably clear dMMR colon tumors in a substantial fraction of patients, the traditional sequence of surgery first, systemic therapy second may need to be reconsidered. Neoadjuvant checkpoint blockade could become the default starting point for newly diagnosed dMMR colon cancer, with surgery reserved for those who show residual disease on imaging, endoscopy, or biomarker testing.
In the most optimistic scenario, a subset of patients with complete clinical and molecular responses might be able to avoid surgery altogether, sparing them the risks of bowel resection, stoma formation, and long-term changes in gastrointestinal function. That kind of organ-preserving approach has already begun to reshape care in rectal cancer, where non-operative management is now an option for carefully selected patients who respond completely to chemoradiation or immunotherapy.
For now, however, the message for patients is one of cautious opportunity rather than immediate change. Neoadjuvant immunotherapy for dMMR colon cancer remains an emerging strategy best pursued in the context of clinical trials or multidisciplinary centers with experience in immuno-oncology and advanced imaging. Surgeons, medical oncologists, radiation oncologists, pathologists, and radiologists will all need to collaborate closely to interpret responses and decide when, or if, an operation can safely be deferred.
As more data accumulate from NICHE-2, the pembrolizumab trial, and future studies that incorporate ctDNA and longer follow-up, the field will move closer to answering the question that now looms over every newly diagnosed dMMR colon cancer: is surgery always necessary, or has immunotherapy finally given some patients a path to cure without the knife?
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
