Can Dual Immunotherapy Replace Surgery in Gastric Cancer?

Dual checkpoint blockade allowed 70.6% of patients with microsatellite instability-high (MSI-H) resectable gastric or gastroesophageal junction adenocarcinoma (G/GEJAC) to avoid surgery in a small cohort of the INFINITY study.

MSI-H tumors account for roughly 10% of early G/GEJACs. They respond well to immunotherapy, with high rates of pathologic complete responses. The Italian INFINITY trial set out to test whether some patients with these tumors might not need gastrectomy.

The trial treated MSI-H patients with durvalumab 1500 mg once a month for 3 months along with 1 300-mg dose of the cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) blocker tremelimumab on day 1. The 18 patients in cohort 1 proceeded to surgery, with a 60% pathologic complete response rate. An additional 18 patients in cohort 2 were the subject of a presentation at the American Association for Cancer Research (AACR) Annual Meeting 2026. These patients were assessed for clinical complete response; if present, they went on to surveillance; if not, they had surgery.

To qualify for a clinical complete response and surveillance, patients were required to have negative findings on CT and PET scans; tumor-informed circulating tumor DNA (ctDNA); and upper endoscopy with ultrasound, including bite-on-bite biopsies and nodal sampling. Surveillance afterward included CT, endoscopy with biopsies, and ctDNA every 12 weeks for up to 2 years.

Among 17 evaluable patients, 1 withdrew consent during immunotherapy, 13 (76%) had a clinical complete response and started surveillance, and the other 4 went to surgery. One patient in the surveillance group had a local regrowth after 4 months, underwent salvage surgery, and remained disease-free. At a median follow-up of 27.1 months, there were no additional progression events.

Overall, 12 of the 17 patients (70.6%) were gastrectomy-free at 2 years without additional treatment. Progression-free survival was 94.1%, and all patients were alive.

“The results are very encouraging,” lead investigator Alberto Leone, MD, said while presenting the results at the AACR annual meeting.

“Nonoperative management could be a safe and effective strategy for patients achieving a clinical complete response after only 3 months of dual immunotherapy,” said Leone, who is a gastrointestinal medical oncologist at the Istituto Nazionale dei Tumori, Milan, Italy. “However, the optimal strategy needs to be established in larger randomized trials.”

Study discussant Yelena Janjagian, MD, gastrointestinal medical oncologist at Memorial Sloan Kettering Cancer Center in New York City, said the findings were important, particularly given that 70.6% of patients avoided a potentially life-altering gastrectomy.

In addition to surgery, the study also calls into question the need for chemotherapy, long the backbone of management alongside surgery, she said. To replace it, however, “it appears that dual checkpoint blockade will be required for a chemotherapy-free approach to achieve organ preservation.”

“Anti-PD-1 alone is not sufficient; we need CTLA-4 to expand and reactivate tumor-specific immunity,” Janjagian continued.

Ultimately, she expects immunotherapy to shift management of MSI-H cancers away from surgery, although some patients will still likely need an operation.

In addition to being MSI-H, patients in the study were mismatch repair deficient and Epstein-Barr virus-negative with T2/T3 tumors; T4 tumors were excluded.

Tumor-agnostic plasma ctDNA was positive at baseline in 13 patients and cleared in 11 after treatment. Higher baseline plasma ctDNA trended toward a lower likelihood of reaching a clinical complete response. Specificity was 100%, so when positive, the test was “very highly informative,” Leone said.

Three patients had grade 3 adverse events (hyperthyroidism, increased gamma-glutamyl transferase, and colitis) that resolved with steroids. There were no grade 4 events, treatment discontinuation, or deaths.

The work was funded by the GONO Foundation and AstraZeneca, the maker of durvalumab and tremelimumab. Leone reported having no disclosures. Janjagian reported having extensive industry ties, including travel funding, consulting fees, and research support from AstraZeneca.

M. Alexander Otto is a physician assistant with a master’s degree in medical science and a journalism degree from Newhouse. He is an award-winning medical journalist who worked for several major news outlets before joining Medscape. Alex is also an MIT Knight Science Journalism fellow. Email: aotto@medscape.net 

A version of this article first appeared on Medscape.com.