Is Non-Surgical Treatment Possible for Stomach Cancer? 62.5% Complete Response with Dostarlimab

Treating Without Removing the Stomach? Dostarlimab "Watch-and-Wait" in dMMR/MSI-H Gastric and Gastro-oesophageal Cancer

After non-surgical treatment with dostarlimab in rectal cancer, the same approach has now been tested systematically for the first time in gastric and gastro-oesophageal junction cancer. In the first interim analysis of the phase 2 GERCOR-GONO DEWI study, presented on October 23, 2026 at the ESMO Congress in Madrid, the tumor disappeared clinically at one year in 15 of 24 patients (62.5%) with localized mismatch repair-deficient (dMMR) or microsatellite instability-high (MSI-H) disease treated with dostarlimab alone, and these patients did not undergo surgery. In a later analysis, the rate reached 67.4% in 43 patients. However, the results are early, and both the definition of complete response and the optimal treatment duration remain to be established. We have previously covered the approval of dostarlimab in dMMR rectal cancer and ESMO's 2026 gastro-oesophageal cancer guideline.

Why does this matter?

Gastric cancer is common in Türkiye and is often diagnosed at an advanced stage. For localized disease, the standard treatment is removal of all or part of the stomach (gastrectomy) together with chemotherapy before and after surgery. This operation has a marked impact on quality of life: weight loss, eating difficulties, vitamin deficiencies and, particularly in older patients, a risk of serious complications.

In about 5–10% of gastric cancers, the tumor's DNA repair system is defective (dMMR/MSI-H). Because these tumors carry a large number of mutations, they are "visible" to the immune system and respond exceptionally well to immunotherapy, while deriving less benefit from conventional chemotherapy. dMMR/MSI-H was one of the first biomarkers to receive a tumor-agnostic approval, and this biology has also changed the role of immunotherapy in colon cancer. In rectal cancer patients with the same biology, dostarlimab has made it possible for the tumor to disappear completely and for surgery to be avoided. The DEWI study asks whether this "organ preservation" approach also applies to the stomach.

What is watch-and-wait in dMMR/MSI-H gastric cancer?

Watch-and-wait means that patients whose tumor has completely disappeared on endoscopy, biopsy and imaging after immunotherapy (clinical complete response) do not undergo surgery and are instead monitored with regular check-ups. In the DEWI study, about two-thirds of patients with localized dMMR/MSI-H gastric and gastro-oesophageal junction cancer treated with dostarlimab achieved a clinical complete response at one year and avoided surgery.

How was the DEWI study conducted?

Results

Of the 25 patients enrolled, one was withdrawn before starting treatment (because of pernicious anemia). Of the 24 patients who received dostarlimab, 23 reached the month-12 assessment, and 14 of them (61%) had a clinical complete response. After the interim analysis, one more patient achieved a clinical complete response at one year, bringing the overall rate to 15 of 24 patients (62.5%). Nine patients proceeded to surgery.

In a later analysis, a clinical complete response with the watch-and-wait approach was achieved in 29 of 43 patients (67.4%). All patients with a complete response at one year already had a response at month 6, meaning that responders become apparent early. At a median follow-up of 14.1 months, none of these patients had a recurrence or loss of life.

What was found in patients who had surgery?

None of the 9 patients who underwent surgery had a pathological complete response; in two-thirds, more than half of the tumor cells were viable. This suggests that patients in whom endoscopy and biopsy did not show a complete response were indeed non-responders or partial responders, in other words, that patient selection was largely accurate. One of the patients who had surgery developed distant metastases.

In the context of other studies

In other immunotherapy studies in localized dMMR/MSI-H gastric cancer, pathological complete response rates after surgery of roughly 40–60% have been reported. In rectal cancer, clinical complete response rates reach 75–100%. The 62.5% clinical complete response rate in DEWI is promising for the stomach; however, how closely a clinical complete response matches a pathological complete response has not yet been proven in the stomach.

Safety

Grade 3–4 treatment-related adverse events occurred in 5 of 24 patients (21%). The most striking finding was that 4 of the 9 patients who underwent surgery developed a fistula (a leak from the surgical suture line), and one patient was lost after surgery due to fistula-related complications. Whether surgery after immunotherapy carries a higher risk of complications needs to be evaluated in more patients.

How should the results be interpreted?

Strengths

• One of the first studies to systematically test non-surgical monitoring in localized dMMR/MSI-H gastric cancer using predefined rules.
• Clinical complete response was based on a strict definition combining endoscopy, biopsy and CT.
• The interim analysis target was comfortably exceeded, and the rate rose further in the later analysis.
• No recurrence has been seen so far in patients with a complete response at one year.

Points of caution

• Small, early data: An interim analysis of 24 patients with a median follow-up of 14 months. In gastric cancer, most recurrences occur within the first 2–3 years; long-term reliability is not yet known.
• No comparison: Single-arm study; although pathological complete response is strongly associated with long-term outcome in other cancers, survival was not compared with standard treatment (surgery + systemic therapy).
• Definition of complete response: Tumor remaining in the deeper layers of the stomach wall or in the lymph nodes can be missed on endoscopic biopsy. A reliable monitoring method such as the one used in the rectum has not yet been established for the stomach; circulating tumor DNA (ctDNA) may not be sufficient to predict response in the stomach.
• Treatment duration: Responders received dostarlimab for up to 24 weeks; whether this is sufficient or whether treatment should continue for up to 2 years is unknown.
• Non-responders: Why some dMMR/MSI-H tumors do not respond, and whether adding a second immunotherapy (anti-CTLA-4) or chemotherapy helps, is unknown.
• Surgical complications: The fistula rate was high among patients who proceeded to surgery.

What does this mean in practice?

• MMR/MSI testing in every gastric cancer: In localized gastric and gastro-oesophageal junction cancer, MMR (immunohistochemistry) or MSI should be tested on the biopsy sample before treatment is planned. This result can fundamentally change treatment.
• If dMMR/MSI-H, consider immunotherapy first: Chemotherapy offers limited benefit in these patients; pre-operative immunotherapy options and clinical trials should be considered.
• Watch-and-wait is not yet standard: DEWI is a phase 2 interim result. Non-surgical monitoring should be considered particularly in older patients, those in poor general condition or at high surgical risk, at experienced centers and preferably within a clinical trial.
• A benefit even if surgery is needed: A good response to immunotherapy may allow a more limited operation and preservation of part of the stomach in some patients.
• Close monitoring is essential: In patients managed without surgery, regular upper gastrointestinal examination and endoscopy, biopsy and imaging must not be missed.
• Hereditary cause: Some dMMR tumors are associated with Lynch syndrome; genetic counseling is recommended, especially in young patients and when there is a family history of colon, uterine or gastric cancer.

Practical advice for patients and families

• If you have been diagnosed with gastric or gastro-oesophageal junction cancer, ask whether your pathology report includes an MMR or MSI result.
• If the result is dMMR/MSI-H, discuss immunotherapy options and clinical trials with your oncologist before surgery.
• If non-surgical monitoring is recommended, do not miss follow-up endoscopies and CT scans; report new symptoms such as difficulty swallowing, bleeding or weight loss right away.
• During immunotherapy, report side effects such as diarrhea, shortness of breath, skin rash or excessive fatigue without delay.
• Pay attention to your nutrition; if you are losing weight, ask for nutritional support.
• If relatives developed colon, uterine or gastric cancer at a young age, seek genetic counseling; also discuss H. pylori testing and treatment (can gastric cancer be prevented?; hereditary factors and H. pylori).

Frequently asked questions

Can gastric cancer be treated without surgery?

The standard treatment is still surgery combined with systemic therapy. However, in localized gastric cancer with defective DNA repair (dMMR/MSI-H), the tumor disappeared completely at one year in about two-thirds of patients who received dostarlimab in the DEWI study, and these patients did not undergo surgery. This approach is still experimental.

What does dMMR/MSI-H mean?

It indicates that the system that repairs DNA copying errors in tumor cells is defective. These tumors carry many mutations and respond very well to immunotherapy. It is found in about 5–10% of gastric cancers and can be detected with a simple stain on the pathology sample.

What happens if the disease comes back during watch-and-wait?

If recurrence is detected early in a patient monitored with regular endoscopy and imaging, surgery usually remains an option. In DEWI, no recurrence was seen at a median follow-up of 14 months in patients with a complete response at one year, but long-term data are not yet available.

Can dostarlimab be used for gastric cancer in Türkiye?

The use of dostarlimab in localized gastric cancer is not covered by its license or by SGK reimbursement; however, it may be possible with off-label approval or within a clinical trial.

Glossary

dMMR / MSI-H

DNA mismatch repair deficiency / microsatellite instability-high; a tumor feature that predicts response to immunotherapy.

Clinical complete response (cCR)

No detectable tumor on endoscopy, biopsy and imaging.

Pathological complete response (pCR)

Absence of viable tumor cells in the tissue removed at surgery.

Watch-and-wait

Monitoring with close check-ups, without surgery, in a patient who has achieved a clinical complete response.

Gastrectomy

Surgical removal of part or all of the stomach.

Becker tumor regression grade (TRG)

A pathology system that grades treatment response according to the proportion of viable tumor remaining in the surgical specimen.

Fistula

An abnormal leak or connection from the suture line after surgery or between organs.

Lynch syndrome

A hereditary condition that can give rise to dMMR tumors and increases the risk of colon, uterine, gastric and other cancers.