Background: The approval of Atezolizumab / Bevacizumab therapy (Atezo/Bev) in 2020 opened up a promising new treatment option for patients with end-stage hepatocellular carcinoma (HCC). However, liver transplant (LTx) patients with HCC are still denied this therapy owing to concerns about ICI-induced organ rejection and lack of regulatory approval.

Methods: A prospective observational study at a tertiary liver transplant centre monitored the compassionate, off-label use of Atezo/Bev in a single, stable LTx recipient with non-resectable HCC recurrence. Close clinical, laboratory and immunological monitoring of the patient was performed throughout a four-cycle Atezo/Bev treatment. Measured parameters were selected after a systematic review of the literature on predictive markers for clinical response and risk of graft rejection caused by ICI therapy.

Results: 19 articles describing 20 unique predictive biomarkers were identified. The most promising negative prognostic factors were the baseline values and dynamic course of IL-6, alpha-fetoprotein (AFP) and the AFP/CRP ratio. The frequency of regulatory T cells (Treg) reportedly correlates with the success of ICI therapy. PD-L1 and CD28 expression level with the allograft, peripheral blood CD4+ T cell numbers and Torque Teno Virus (TTV) titre may predict risk of LTx rejection following ICI therapy. No relevant side effects or acute rejection occurred during Atezo/Bev therapy; however, treatment did not prevent tumor progression. Absence of PD-L1 expression in pre-treatment liver biopsies, as well as a progressive downregulation of CD28 expression by CD4+ T cells during therapy, correctly predicted absence of rejection. Furthermore, increased IL-6 and AFP levels after starting therapy, as well as a reduction in blood Treg frequency, correctly anticipated a lack of therapeutic response.

Conclusion: Atezo/Bev therapy for unresectable HCC in stable LTx patients remains a controversial strategy because it carries a high-risk of rejection and therapeutic response rates are poorly defined. Although previously described biomarkers of rejection risk and therapeutic response agreed with clinical outcomes in the described case, these immunological parameters are difficult to reliably interpret. Clearly, there is an important unmet need for standardized assays and clinically validated cut-offs before we use these biomarkers to guide treatment decisions for our patients.