Preventing NK cell activation in the damaged liver induced by cabozantinib/PD-1 blockade increases survival in hepatocellular carcinoma models

Satoru Morita,Hiroto Kikuchi,Grace Birch,Aya Matsui,Atsuyo Morita,Tatsuya Kobayashi,Zhiping Ruan,Peigen Huang,Alexei Hernandez,Erin M Coyne,Sarah M Shin,Mark Yarchoan,Mari Mino-Kenudson,Rizwan Romee,Won Jin Ho,Dan G Duda
DOI: https://doi.org/10.1101/2023.10.20.563378
2023-10-23
bioRxiv
Abstract:The addition of anti-VEGF antibody treatment to immune checkpoint blockade (ICB) has increased the efficacy of immunotherapy in advanced hepatocellular carcinoma (HCC). Despite an initial promise, adding multitargeted kinase inhibitors of VEGFR with ICB has failed to increase survival in HCC. To reveal the mechanisms underlying treatment failure, we studied the effects of cabozantinib/ICB using orthotopic murine HCC models with or without liver damage. We monitored tumor growth and liver function, recorded survival outcomes, and performed immune profiling studies for intra-tumoral and surrounding liver. Cabozantinib/ICB treatment led to tumor regression and significantly improved survival in mice with normal livers. However, consistent with the clinical findings, combination therapy failed to show survival benefits despite similar tumor control when tested in the same models but in mice with liver fibrosis. Moreover, preclinical and clinical data converged, showing that activating immune responses by cabozantinib/ICB treatment induced hepatoxicity. Immune profiling revealed that combination therapy effectively reprogrammed the tumor immune microenvironment and increased NK cell infiltration and activation in the damaged liver tissue. Surprisingly, systemic depletion of NK reduced hepatotoxicity elicited by the combination therapy without compromising its anti-cancer effect, and significantly enhanced the survival benefit even in mice with HCC and underlying liver fibrosis. These findings demonstrate that preventing NK activation allowed for maintaining a favorable therapeutic ratio when combining ICB with cabozantinib in advanced HCC models.
What problem does this paper attempt to address?