The predictive effects of liver metastases for immune-checkpoint inhibitors (ICIs) and the relationship between tumor mutational burden (TMB) and liver metastases (LM) remain unclear.
A systematic review and meta-analysis were conducted to explore the heterogeneity of ICIs efficacy between patients with or without LM. A pan-cancer cohort of 1,661 patients who received ICIs was downloaded and analyzed to assess the association between TMB and LM.
Of 21053 studies identified in our search, eight single-arm studies and 24 randomized controlled trials were included. Overall, 17957 patients with advanced or metastatic cancers (4805 patients (26.8%) with LM and 13151 patients (73.2%) without LM) were enrolled. The pooled objective response rate (ORR) was 8.5% (95% CI 4%–13%) in the LM group versus 21% (95% CI 16%–21%) in the non-LM group. The pooled hazard ratio (HR) for death was 0.85 (95% CI 0.80–0.90) in the LM group treated with ICIs compared with the standard of care. In patients without LM who were treated with ICIs, the pooled HR for death was 0.78 (95% CI 0.73–0.82) compared with the standard of care. The difference in efficacy between patients with or without LM treated with ICIs was significant (p=0.04). Pan-cancer analysis revealed that the TMB-high rate was 10.8% in liver metastatic lesions versus 21.4% in other metastatic lesions (p=0.004). In addition, TMB was also significantly associated with OS as a binary cutoff (p=0.05) and was an independent prognostic variable (HR=0.98, P=0.047) as a continuous variable in patients with LM.
In patients with LM, the efficacy of immunotherapy was attenuated, but TMB-high could predict better survival outcomes.