Dostarlimab has generated extraordinary excitement in colorectal oncology, but much of the coverage conflates different stages and subtypes of the disease. For stage 4 (metastatic) colon cancer, dostarlimab can produce meaningful and sometimes durable responses, though only in a specific molecular subset: tumors that are mismatch repair deficient (dMMR) or have high microsatellite instability (MSI-H). That subset accounts for a strikingly small share of stage 4 cases, roughly 7% of all metastatic colorectal cancers. The widely reported 100% complete response rate came from a trial in locally advanced rectal cancer, a very different clinical scenario from metastatic disease spread to distant organs.
Only a Small Fraction of Stage 4 Cases Are Eligible
Dostarlimab is a PD-1 inhibitor, a type of immunotherapy that works by removing a brake on the immune system so T cells can attack cancer cells more effectively. But this approach only works well when tumors generate enough abnormal proteins for the immune system to recognize. That happens when a tumor’s DNA repair machinery is broken, a condition called mismatch repair deficiency. These dMMR tumors accumulate far more mutations than their repair-proficient counterparts, producing a wealth of targets that the immune system can latch onto once the PD-1 brake is released.
A large U.S. study analyzing over 100,000 colorectal cancers found that about 14% of all colorectal adenocarcinomas were MSI-H or dMMR. But the proportion drops sharply as the disease advances: roughly 17% of stage I cancers, about 20% of stage II, around 12% of stage III, and just 7.3% of stage IV cancers carry this feature.1JCO Precision Oncology. The Prevalence and Prognosis of Microsatellite Instability-High/Mismatch Repair-Deficient Colorectal Adenocarcinomas in the United States That means the vast majority of people diagnosed with stage 4 colon cancer will not be candidates for dostarlimab. For those who are, though, the drug offers something that standard chemotherapy often cannot: the possibility of a deep, lasting response.
What the Evidence Shows for Advanced Disease
The largest body of evidence for dostarlimab in advanced, previously treated dMMR tumors comes from the GARNET trial, a multi-cohort study that enrolled 327 patients with various dMMR solid tumors, including colorectal cancer. Among those who responded to treatment, the results were encouraging: the median duration of response had not yet been reached even after years of follow-up, meaning many responders were still benefiting from the drug when the data were analyzed. About 72% of responders maintained their response for at least a year, and the probability of still being in response at two years was roughly 85%.2JAMA Network Open. Antitumor Activity and Safety of Dostarlimab Monotherapy in Patients With Mismatch Repair Deficient Solid Tumors
Those numbers, while genuinely impressive for advanced cancer, also reveal a more complex picture than the headlines convey. Across all dMMR solid tumors in the study, the median progression-free survival was about seven months. That means roughly half the patients saw their disease advance within that window, while the other half had longer benefit. The probabilities of remaining progression-free at one, two, and three years were approximately 46%, 41%, and 40%, respectively.3JAMA Network Open. Antitumor Activity and Safety of Dostarlimab Monotherapy in Patients With Mismatch Repair Deficient Solid Tumors In other words, a sizeable minority of patients with advanced dMMR disease got long-lasting benefit that looked almost like a plateau, while others progressed despite the drug. This is a pattern seen across PD-1 inhibitors in various cancers: when immunotherapy works, it tends to work for a long time, but it does not work for everyone even within the “right” molecular subgroup.
Why the 100% Response Rate Doesn’t Apply to Stage 4
The result that catapulted dostarlimab into mainstream news was a trial at Memorial Sloan Kettering Cancer Center in which every single patient with locally advanced dMMR rectal cancer achieved a clinical complete response after six months of treatment. The initial report covered 12 patients: all showed no detectable tumor on imaging, endoscopy, or biopsy, and none required surgery, radiation, or chemotherapy.4PubMed Central. PD-1 Blockade in Mismatch Repair-Deficient, Locally Advanced Rectal Cancer Updated data expanded the cohort to 41 patients, with the same outcome: every patient responded completely, and with a median follow-up of nearly 29 months, no one had experienced recurrence or needed additional treatment.5Journal of Clinical Oncology. Durable complete responses to PD-1 blockade alone in mismatch repair deficient locally advanced rectal cancer
These are remarkable results by any measure. But they came from a fundamentally different patient population than someone with stage 4 colon cancer. “Locally advanced” means the tumor is large or has grown into nearby tissue, but it has not spread to distant organs like the liver or lungs. That is a critical distinction. A locally advanced tumor confined to one area is a very different challenge for the immune system than metastatic disease that has seeded multiple sites throughout the body. The tumor burden is lower, the immune microenvironment may be more favorable, and the cancer has had less time to evolve escape mechanisms. Extrapolating a 100% complete response rate from locally advanced rectal cancer to stage 4 metastatic colon cancer would be a mistake, and no oncologist is doing so.
How Dostarlimab Compares to Other Immunotherapies in This Space
Dostarlimab is not the only PD-1 inhibitor used for dMMR colorectal cancer. Pembrolizumab and nivolumab have both been studied extensively and are widely used. Some comparative analyses suggest dostarlimab may have superior response rates in certain colorectal cancer subgroups, though head-to-head randomized trials between these drugs are limited.6Current Cancer Therapy Reviews. Dostarlimab Unveiled: Breakthroughs in Colorectal Cancer Related to Genomic Aberration Coupled with ncRNA Functions In practice, the choice between these agents often depends on the clinical setting, available trial data for the specific indication, and regulatory approvals rather than large differences in how well they work. The broader point is that dMMR status is the key predictor of benefit, more so than which specific PD-1 inhibitor is used.
The underlying biology explains why. Tumors with broken DNA repair systems accumulate mutations at a high rate, and many of these mutations produce abnormal proteins on the cell surface. The heavier this mutational load, the more targets the immune system has, and the more likely a PD-1 inhibitor is to trigger a robust anti-tumor response. Research has shown that the specific type of mutations matters too: dMMR tumors disproportionately accumulate insertions and deletions rather than simple single-letter changes, and these insertions and deletions tend to produce especially foreign-looking proteins that the immune system recognizes more readily.7PubMed Central. Dostarlimab in the treatment of mismatch repair deficient recurrent or advanced endometrial cancer
Side Effects in Context
One of the reasons immunotherapy has been embraced so enthusiastically is that its side-effect profile tends to be more tolerable than traditional chemotherapy. In the GARNET trial, the majority of treatment-related side effects were mild, with about three-quarters classified as low-grade. The most common were fatigue (about 18% of patients), diarrhea (about 14%), and nausea (about 14%). Severe side effects occurred in roughly 17% of patients, and about 6% discontinued treatment because of them. No deaths in the trial were attributed to the drug itself.8Journal for ImmunoTherapy of Cancer. Safety and antitumor activity of dostarlimab in patients with advanced or recurrent DNA mismatch repair deficient/microsatellite instability-high or proficient/stable endometrial cancer
That said, PD-1 inhibitors carry a unique class of risks that chemotherapy does not. By releasing the immune system’s brakes, they can occasionally trigger the immune system to attack healthy organs. These immune-related events can affect the thyroid, liver, lungs, colon, or other tissues. Most are manageable with steroids or treatment interruption, but some can be serious. Your oncology team will monitor bloodwork and symptoms throughout treatment. For someone with stage 4 disease weighing dostarlimab against further chemotherapy, the side-effect comparison generally favors immunotherapy in terms of day-to-day quality of life, but the immune-related risks are distinct and worth understanding.
Quality of Life During Treatment
For patients with advanced cancer, how a treatment affects daily life matters as much as tumor shrinkage. Patient-reported outcome data from the GARNET trial in the dMMR colorectal cancer subgroup showed that quality of life and most gastrointestinal symptoms improved from baseline and remained stable or better from cycle two through cycle seven. Diarrhea symptoms, a common concern in colorectal cancer, remained stable throughout treatment.9Journal of Clinical Oncology. Patient-reported outcomes from the GARNET trial in patients with advanced or recurrent mismatch repair deficient colorectal cancer In the context of advanced disease, where many treatments gradually erode how well a person feels, maintaining or improving quality of life over months of therapy is a meaningful outcome on its own.
Getting Tested for dMMR/MSI-H Status
Because dostarlimab only benefits the dMMR/MSI-H subset, testing is the gateway to the entire conversation. Two standard methods are used: immunohistochemistry (IHC), which looks directly at whether the repair proteins are present in tumor tissue, and a PCR-based test that checks for the instability pattern in the DNA itself. Both are considered equally valid as initial screening methods for dMMR/MSI-H in colorectal cancer.10PubMed. MSI/MMR-deficient tumor diagnosis: Which standard for screening and for diagnosis? Most major cancer centers now perform one or both of these tests routinely on all new colorectal cancer diagnoses, though testing rates are not universal.
Newer genomic sequencing approaches (next-generation sequencing, or NGS) can also detect MSI, but their accuracy is not perfect. One study found that a commonly used NGS-based tool missed the MSI signal in roughly 10-16% of confirmed dMMR metastatic colorectal cancers when run on standard gene panels, and some of the missed cases included patients who actually responded to immunotherapy.11PubMed. Performance of Next-Generation Sequencing for the Detection of Microsatellite Instability in Colorectal Cancer With Deficient DNA Mismatch Repair If you have stage 4 colon cancer and your NGS report says MSI-stable, but there is any clinical suspicion of dMMR (family history of Lynch syndrome, for instance, or a right-sided tumor with certain pathologic features), it may be worth confirming with IHC or PCR rather than relying solely on the genomic panel.
Disparities in Who Gets Identified
The prevalence of dMMR/MSI-H colorectal cancer is not evenly distributed across racial and ethnic groups. In the same large U.S. study that established the stage-by-stage breakdown, non-Hispanic White patients had the highest rate of dMMR/MSI-H tumors at about 17%, while non-Hispanic Black patients had the lowest at about 11%. Hispanic patients fell in between at roughly 15%, and Asian/Pacific Islander patients at about 12%.12JCO Precision Oncology. The Prevalence and Prognosis of Microsatellite Instability-High/Mismatch Repair-Deficient Colorectal Adenocarcinomas in the United States
These differences in prevalence mean that, at the population level, a smaller share of stage 4 patients in some racial groups will be eligible for PD-1 inhibitors like dostarlimab. But the more pressing equity concern is whether all patients are being tested in the first place. Guidelines now recommend universal dMMR/MSI testing for all colorectal cancers, regardless of age or family history, yet adherence to universal testing varies widely by institution and region. A patient whose tumor never gets tested for dMMR will never be offered an immunotherapy that could change their trajectory.
Ongoing Trials and Where the Field Is Headed
Most existing dostarlimab data in colon cancer come from either the locally advanced rectal cancer studies or from the basket-trial approach of GARNET, which enrolled various dMMR solid tumors rather than focusing specifically on colon cancer by stage. The field is now moving toward more targeted questions. AZUR-2 is a large Phase III randomized trial testing dostarlimab before and after surgery in patients with resectable (operable) dMMR/MSI-H colon cancer, specifically those with T4N0 or stage III disease. About 711 patients will be randomized to receive dostarlimab around the time of surgery versus standard care. The primary endpoint is event-free survival.13PubMed Central. A Phase III study of perioperative dostarlimab in patients with dMMR/MSI-H resectable colon cancer: AZUR-2 study design
AZUR-2 is focused on earlier-stage, operable disease rather than stage 4, but its results will shape how dostarlimab is used across the colon cancer treatment landscape. If perioperative immunotherapy proves superior to chemotherapy or surveillance in these patients, it will reinforce the broader shift toward making dMMR status, rather than just anatomic stage, the central factor in treatment planning. For stage 4 specifically, the open question is whether dostarlimab combined with other agents, such as chemotherapy or other immunotherapy drugs, can push response rates higher than monotherapy achieves. Several trials exploring such combinations are in various stages of enrollment.
The Cost of Treatment
Immunotherapy drugs are expensive, and dostarlimab is no exception. While no published cost-effectiveness analysis focuses specifically on dMMR stage 4 colon cancer, a study modeling dostarlimab plus chemotherapy versus chemotherapy alone in dMMR endometrial cancer found that the combination produced a substantial survival benefit (about 5.5 additional quality-adjusted life-years) but at a high incremental cost, yielding a cost-effectiveness ratio of roughly $60,000 per quality-adjusted life-year.14PubMed Central. Cost-effectiveness of dostarlimab plus chemotherapy for primary advanced or recurrent endometrial cancer That figure is considered favorable by conventional thresholds in the U.S. and several other countries, though it was specific to the dMMR endometrial cancer population and the combination regimen studied. In non-dMMR disease, the same analysis found the economics were far less favorable, underscoring once again that the molecular subtype drives everything, including value.
For an individual patient with stage 4 dMMR colon cancer, the practical cost picture depends heavily on insurance coverage, access to patient assistance programs, and whether treatment is received at a center participating in a clinical trial (where the drug is often provided at no cost). If you are exploring dostarlimab, asking your oncology team about open trials is worth the conversation, both for potential access to the drug and for contributing to the evidence base that will help future patients.
When Immunotherapy Does Not Work in dMMR Tumors
Even within the dMMR/MSI-H population, not everyone responds. Researchers are working to understand why some dMMR tumors resist PD-1 blockade, and the intensity of microsatellite instability appears to be one factor. Not all dMMR tumors are equally unstable; some have a heavier mutational burden than others, and tumors with more intense instability seem to respond more consistently. It has been proposed that the degree of MSI, rather than a simple present-or-absent classification, could eventually help predict who will and won’t benefit from anti-PD-1 therapy.15PubMed Central. Dostarlimab in the treatment of mismatch repair deficient recurrent or advanced endometrial cancer
Other potential resistance mechanisms include loss of the molecular pathways that cancer cells need to display their abnormal proteins to the immune system, an immunosuppressive local environment around the tumor, or acquired mutations that help the cancer evade immune detection after an initial response. These are active areas of research without definitive clinical tools yet. For now, the practical takeaway is that dMMR status is necessary but not sufficient: it opens the door to immunotherapy, but it does not guarantee that the door leads somewhere. Oncologists managing stage 4 disease have to weigh the meaningful chance of a durable response against the possibility of progression and the need for backup treatment plans.