Background
Phase II single-arm trial. 115 patients with dMMR (mismatch repair-deficient) non-metastatic colon cancer (stage III or high-risk stage II), eligible for curative-intent surgery. Assessed feasibility and efficacy of short-course neoadjuvant nivolumab + ipilimumab followed by planned resection, building on the NICHE-1 pilot.
Interventions and follow up
Regimen: Nivolumab 240 mg IV (week 1) + ipilimumab 1 mg/kg IV (week 1), then nivolumab 240 mg IV (week 3) → surgery at week 5–6
Design: Single-arm; no comparator. All patients proceeded to planned resection
Primary endpoint: (1) absence of progression preventing surgery; (2) pathological complete response (pCR, ypT0N0)
Median follow-up: 26.1 months
Design: Single-arm; no comparator. All patients proceeded to planned resection
Primary endpoint: (1) absence of progression preventing surgery; (2) pathological complete response (pCR, ypT0N0)
Median follow-up: 26.1 months
Results
Feasibility (surgery as planned): 113/115 (98.3%)
Major pathological response (≤10% residual): 105/111 (95%)
pCR (0% residual, ypT0N0): 75/111 (68%)
Recurrence-free survival: 0 recurrences at median 26 months
Major pathological response (≤10% residual): 105/111 (95%)
pCR (0% residual, ypT0N0): 75/111 (68%)
Recurrence-free survival: 0 recurrences at median 26 months
Adverse events
Grade ≥3 immune-related AEs: 5 patients (4%)
Any-grade irAEs: fatigue, diarrhea, rash, elevated liver enzymes; no treatment-related deaths
Surgical impact: no progression preventing surgery; 2 non-progression-related delays
Any-grade irAEs: fatigue, diarrhea, rash, elevated liver enzymes; no treatment-related deaths
Surgical impact: no progression preventing surgery; 2 non-progression-related delays
Conclusions
Short-course neoadjuvant nivolumab + ipilimumab produced major pathological response in 95% and pCR in 68% of dMMR colon cancer patients, without compromising surgical feasibility (98.3% underwent timely surgery). No recurrences were observed at median 26 months, suggesting deep and potentially durable disease eradication.
Key Limitations
Single-arm phase II without randomized comparator; no comparison with surgery alone or single-agent PD-1 blockade. Median follow-up of 26 months is insufficient to assess long-term DFS/OS. Exclusively dMMR population limits generalizability (∼15% of colorectal cancers). The 32% not achieving pCR still required complete resection; criteria for non-operative management not yet defined. Randomized evidence pending from NICHE-3 and NICHE-4.
Clinical Context
NICHE-2 is the largest prospective study demonstrating the sensitivity of dMMR non-metastatic colon cancer to short-course neoadjuvant dual checkpoint blockade. The 68% pCR and 95% major pathological response have catalyzed randomized trials (NICHE-3: nivo+ipi vs surgery; NICHE-4: non-operative management in pCR responders). Neoadjuvant nivo+ipi is not yet FDA-approved for this setting; ESMO guidance is evolving as practice shifts toward neoadjuvant immunotherapy for operable dMMR colon cancer at specialized centers.