Short-course neoadjuvant immunotherapy produced durable long-term outcomes in patients with locally advanced mismatch repair-deficient (dMMR) colon cancer, with 3-year disease-free survival and 5-year overall survival both reaching 100%, according to long-term results from the phase 2 NICHE-2 trial published in The Lancet.
The findings extend previously reported NICHE-2 results showing high pathological response rates after just four weeks of neoadjuvant nivolumab plus ipilimumab.
“Immunotherapy is changing the treatment paradigm and should be considered early in multidisciplinary discussions for appropriately selected patients with locally advanced dMMR colon cancer,” Daryl R. Ramai, MD, MPH, MSc, director of third space endoscopy and assistant professor of medicine at the University of Utah School of Medicine, who was not involved in the study, told GI & Hepatology News.
Long-term outcomes after short-course therapy
NICHE-2, led by Myriam Chalabi, MD, PhD, of the Netherlands Cancer Institute in Amsterdam, was a single-arm, multicenter phase 2 trial conducted at six hospitals in the Netherlands. Adults with previously untreated, nonmetastatic, resectable clinical stage 2 or 3 dMMR colon cancer received one dose of ipilimumab at 1 mg/kg on day 1 and two doses of nivolumab at 3 mg/kg on days 1 and 15, followed by surgery. The primary efficacy endpoint was 3-year disease-free survival, measured from surgery. With no control arm, the investigators prespecified a 93% rate as success, against a historical rate of 82%.
Among 115 treated patients, 111 were included in the per-protocol efficacy analysis. Across all 115, the median age was 60 years, 58% were women, and 32% had Lynch syndrome. At baseline, 67% were clinically node-positive, and 64% had cT4 tumors. All 111 patients completed both cycles of immunotherapy and underwent surgery, with microscopically clear (R0) margins achieved in every patient.
A pathological response was observed in 98% of patients, including a major pathological response in 95% and a pathological complete response in 68%. Three patients with tumor-positive lymph nodes received adjuvant chemotherapy.
At a median postoperative follow-up of 58.7 months, all patients in the per-protocol population were alive and disease-free, resulting in a 3-year disease-free survival rate of 100% and meeting the primary efficacy endpoint (P < .0001). In an exploratory post hoc analysis, 5-year overall survival was also 100% after a median follow-up of 60.4 months from study enrollment. In a post hoc intention-to-treat analysis of all 115 treated patients, 3-year disease-free survival was 99%, and 5-year overall survival was 100%.
“The updated analysis provides important evidence that the impressive pathological responses translate into durable disease control,” Dr. Ramai said. “These findings strengthen the significance of the original results and provide reassurance regarding the long-term efficacy of this short-course immunotherapy approach.”
The outcomes were observed in a population enriched for high-risk disease. Patients with high-risk stage 2 or stage 3 dMMR colon cancer remain at a 20% to 40% risk of recurrence after current standard treatment, the authors noted.
Circulating tumor DNA
The study also evaluated circulating tumor DNA (ctDNA), measured with the tumor-informed Signatera assay (Natera), as a potential marker of treatment response and minimal residual disease.
ctDNA was detectable at baseline in 99 of 108 evaluable patients, or 92%. Among patients with detectable ctDNA at baseline, clearance occurred in 45% after the first immunotherapy cycle and in 82% before surgery after the second cycle.
Clearance before surgery varied by pathological response. It occurred in 92% of patients who ultimately achieved a pathological complete response and 70% of those with a near-complete pathological response.
At the postoperative minimal residual disease assessment, all patients were ctDNA-negative.
Presurgical ctDNA positivity had 93% specificity for residual pathological disease but only 35% sensitivity, the authors reported. Dr. Ramai added that “six patients with negative ctDNA still had viable lymph node metastases.”
The authors wrote that ctDNA may eventually help with risk stratification but requires prospective validation before it can be used to guide treatment decisions.
Organ preservation
The high pathological complete response rate also raises the possibility that some patients with dMMR colon cancer may ultimately be candidates for organ-preserving approaches, the authors wrote.
For now, however, surgery remains the standard of care for localized dMMR colon cancer, and the investigators emphasized that further study is needed before nonoperative management can be considered.
“Despite the remarkable outcomes, all patients in the per-protocol population underwent surgery, meaning the study does not establish the safety of nonoperative management,” Dr. Ramai said. “Ultimately, carefully selected patients may be candidates for organ-preserving strategies, but randomized controlled studies are needed to establish the safety and oncological efficacy of nonoperative management before colectomy can be safely omitted.”
Endoscopic assessment of complete response is not yet standardized in colon cancer, according to the authors. Ongoing NICHE cohorts are testing whether longer treatment and a longer interval to surgery increase pathological complete response rates and whether endoscopy can predict pathological response; the authors wrote that the results will inform future organ-preservation studies.
The authors acknowledged that the single-arm design, with no control group, precludes direct comparison with standard treatment. They also wrote that because eligibility relied on radiographic staging, a neoadjuvant approach risks overtreating some patients with low-risk stage 2 tumors. Results from randomized studies, including the phase 3 AZUR-2 trial of perioperative anti-PD-1 therapy versus standard adjuvant chemotherapy, are awaited. Dr. Ramai similarly noted that further randomized studies are needed to establish comparative efficacy, optimize treatment duration, and clarify the role of adjuvant therapy.
Despite those limitations, the investigators concluded that the long-term results support neoadjuvant immune checkpoint blockade as a highly effective treatment strategy and that it could be considered as part of the standard of care for patients with locally advanced dMMR colon cancer.
The study was funded by Bristol Myers Squibb. Dr. Chalabi reported advisory roles with Bristol Myers Squibb, Kineta, Merck Sharp & Dohme, NOUSCOM, and Roche/Genentech, as well as institutional research grants unrelated to this study from Bristol Myers Squibb, Agenus, Merck Sharp & Dohme, and Roche/Genentech. Two co-authors are affiliated with and hold stock in Natera, which makes the ctDNA assay used in the trial; other co-authors reported ties to companies including AstraZeneca, Novartis, and Sanofi.