Immunotherapy in patients with early-stage microsatellite instable (MSI-H)/mismatch repair deficient (dMMR) colorectal cancer (CRC) treated in a community setting.

e15668 Background: MSI-H/dMMR cancers represent 10-15% of early-stage colorectal cancer (CRC) and respond well to immune checkpoint therapy. However, clinical trials often exclude frail, elderly patients with an ECOG Performance Status of 2 or higher, focusing instead on younger, healthier individuals. This study presents a single-center analysis of upfront IO in early-stage dMMR/MSI-H CRC, particularly focusing on frail, elderly patients. Methods: We conducted an observational study of patients with early-stage dMMR/MSI-H CRCs treated with IO as either neoadjuvant or definitive therapy, considered unsuitable for upfront surgery due to age and/or comorbidities. We collected and analyzed clinicopathological characteristics and safety data. Results: In this cohort of 19 patients treated with neoadjuvant or definitive IO using either Ipilimumab/Nivolumab or Pembrolizumab between November 2019 and June 2024, the median age was 76.4 years, with 11 (58%) being over 65, 8 (42%) over 80, and 4 (21%) over 90. All patients had early-stage locally advanced disease, with 73% right-sided tumors, 94% sporadic MSI-H, and 78% BRAF mutated; only one had a germline mutation. Notably, 92% of patients had an ECOG PS of 1-2 and an average of 3 comorbidities. In the neoadjuvant cohort (10 patients; 52%), treatment included Pembrolizumab (6/10) or Ipilimumab/Nivolumab (4/10), with all patients proceeding to surgery within a median of 6 weeks. A pCR was achieved in 7/10 (70%), while the remaining 3/10 (30%) showed a partial response. In the definitive therapy group (9 patients; 47%), the median age was 80 years (range 52-90 years). Complete response was observed in 6/9 (66%) by imaging and confirmed in two patients by colonoscopy. As of January 2025, the median duration of response was 10 months, with ongoing follow-up. Treatment was well-tolerated, with 5 patients experiencing immune-related adverse events, including a grade 4 myasthenia gravis case and two with grades 1-2 myalgias and thyroiditis. Baseline circulating tumor DNA (ctDNA) data were available for 7 patients, with 6 initially ctDNA positive; 5 achieved negativity within 6 weeks of treatment. One patient showed rising ctDNA levels post-treatment, leading to a change in therapy. Conclusions: Immunotherapy as a neoadjuvant or definitive treatment is feasible, effective, and safe. It demonstrates high response rates even among older patients with multiple comorbidities.