Abstract

Background:

The mechanisms leading to cancer progression of hepatocellular carcinoma (HCC) such as acquired resistance to immunotherapies and metastatic progression are not clear. Circulating tumor cells (CTCs) have been defined as cancer cells released into the blood circulation from primary tumor. CTCs have notable advantages in that they are noninvasive and real-time biomarker that provide information about metastatic process. CTC analysis has the potential to reveal the mechanism of disease progression and is expected to be applied clinical use such as treatment decision. We investigated the changes in CTC counts and gene expression related to cancer progression in patients with unresectable HCC treated by Atezolizumab plus Bevacizumab.

Methods:

We obtained 5mL of peripheral blood from 15 HCC patients treated with Atezolizumab plus Bevacizumab at baseline and response evaluation. Treatment response was assessed by modified RECIST. CTCs were isolated with RosetteSep^TM Human CD45 Depletion Cocktail and enriched cells were stained with monoclonal antibodies targeting the cell surface antigens including CD45, CD90, CD133, Pan-CK, EpCAM, and Vimentin. The cells were counted by flow cytometry using a FACSAria II. CD45 negative and Pan-CK positive cells were defined as CTCs. RNA was extracted from enriched cells and 373 genes related to cancer progression were investigated by next-generation sequencing (NGS) on Illumina Miseq and gene set enrichment analysis (GSEA) was performed.

Results:

Median CTC counts of PR/SD group decreased at response evaluation, compared with baseline (127 ± 43 vs. 58 ± 6, p < 0.05) and is lower compared with PD group at response evaluation (58 ± 6 vs. 153 ± 12, p < 0.05). In NGS analysis of 373 genes, 99 genes showed significant expression changes in clinical course. Unsupervised hierarchical clustering analysis with the changes of 99 genes expression levels classified into two clusters A and B. Patients in cluster A were responder with Atezolizumab plus Bevacizumab and showed 100.0% survival rate at 1-year, which was better than 50.0% survival rate in cluster B (p = 0.06). GSEA showed that the genes expression of apoptosis signaling pathway (FAS, BCL2) were significantly upregulated in responder group (cluster A). TGF-beta signaling pathway-related genes (TGF-β1, SMAD2) were upregulated in non-responder group (cluster B) in clinical course of Atezolizumab plus Bevacizumab (p < 0.05).

Conclusion:

In HCC patients treated by Atezolizumab plus Bevacizumab, the change of CTC counts was related to therapeutic effect. The gene expression analysis of CTCs showed that apoptosis signaling pathway-related genes were upregulated in responder group and the activation of TGF-beta signaling pathway may play an important role in resistance to Atezolizumab plus Bevacizumab. CTC in HCC patients treated by immunotherapy may be a notable biomarker that reveal molecular signature of cancer progression.