Overall survival of immunotherapy versus standard of care in chemorefractory microsatellite stable metastatic colorectal cancer: a propensity score matched analysis of 708 patients.

Immune checkpoint inhibitors (ICIs) have limited efficacy in proficient mismatch repair/microsatellite stable (pMMR/MSS) metastatic colorectal cancer (mCRC). However, selected patients with specific metastatic patterns may derive benefit. Patients with chemorefractory pMMR/MSS mCRC treated with ICI-based regimens were retrospectively identified. A comparison cohort treated with trifluridine/tipiracil±bevacizumab, regorafenib, or fruquintinib as standard of care (SOC) was generated through 1:1 propensity score matching by age, sex, Eastern Cooperative Oncology Group performance status (ECOG PS), liver metastases (present/absent), and RAS/BRAF status. Overall survival (OS) was compared using Cox regression. A total of 354 patients treated with ICIs and 354 treated with SOC were matched. Median age was 55 years, 52% male, 32% ECOG PS 0, 30% right-sided, and 69% RAS mutated in both groups, while 61% and 60% had liver metastases, respectively. Median OS (mOS) was 10.8 months with ICIs and 9.0 months with SOC (HR 0.76, 95% CI 0.64 to 0.92, p=0.004). In patients without liver metastases, mOS was longer with ICIs than SOC (19.1 vs 13.2 months, HR 0.59, 95% CI 0.43 to 0.80, p<0.001), whereas outcomes were similar in patients with liver metastases (6.4 vs 6.5 months, p=0.303). In univariable analyses, age, sex, primary tumor site, and RAS/BRAF status were not associated with OS. Treatment with ICIs, absence of liver metastases, one prior line of therapy, less than three metastatic sites, and ECOG PS 0 were associated with the most favorable outcomes in univariable and multivariable models. In chemorefractory pMMR/MSS mCRC without liver metastases, ICI-based regimens yielded longer OS than SOC. Further investigation of ICIs in this patient population is warranted.