A new perspective on the immune escape mechanism in HCC: onco-foetal reprogramming

Summary

We found a shared immunosuppressive microenvironment between foetal liver and hepatocellular carcinoma (HCC) which includes the re-emergence of foetal-associated endothelial cells (PLVAP/VEGFR2) and foetal-like (FOLR2) tumour-associated macrophages in HCC, mediated via VEGF–NOTCH signalling. The discoveries suggest possible novel targets for therapeutic interventions in HCC.

Main

Hepatocellular carcinoma (HCC) is the 6th most common cancer worldwide but has the second highest mortality and thus among the common cancers HCC has one of the poorest prognosis.¹ Early diagnosis remains a challenge and only approximately 20% of HCC are amendable to potentially curative ablative therapies when discovered. In addition, current systemic therapies in HCC are poorly efficacious. Known major driver mutations in HCC such as TP53 and CTNNB1 are not easily druggable. An additional hurdle to the development of efficacious therapies and the determination of predictive clinical biomarkers is the presence of extensive intra-tumour heterogeneity (ITH) in HCC, both within the tumour and the tumour microenvironment (TME).² Current first-line therapies for advanced HCC, namely tyrosine-kinase inhibitors lenvatinib and sorafenib, confer overall objective response rates (ORR) of 24% and 9.2% and median OS of 13.6 months and 12.3 months respectively³ and the best ORR of 27% is from combination immunotherapy atezolizumab (anti-PD-L1) plus bevacizumab (anti-VEGF).⁴ There are currently no validated predictive biomarkers.

Existing approaches in HCC research have been mainly based on the bulk sequencing of single HCC biopsies which under-represent the heterogeneous genomics landscape of HCC and grossly classify HCC into molecular subtypes.² Moreover, bulk sequencing also averages out signals from the diverse cell types in the tumour, limiting reconstruction and in-depth examination of the TME. Specifically, for immune cells, bulk tissue analyses do not provide information on the composition, proportion, heterogeneity or spatial distribution that exists in the TME, which are crucial to understanding treatment outcomes. In order to overcome the above limitations, we employed multi-region sampling, single cell RNA sequencing and spatial transcriptomics to demarcate the heterogeneous tumour landscape in HCC.⁵ This enabled comprehensive characterisation of the transcriptomics, spatial, and functional landscape of the HCC tumour ecosystem and the unique cell–cell interactions underlying ITH. Using this approach, we have discovered a previously unexplored immune-suppressive onco-foetal reprogramming of the tumour ecosystem via VEGF–NOTCH signalling in HCC. The implications of such findings to the development of HCC therapeutic strategies are profound.

The VEGF signalling pathway promotes tumour angiogenesis through the activation of a kinase cascade that includes RAS and mitogen-activated protein kinase (MAPK). Beyond this role, VEGF harbours immunomodulatory effects in the TME.⁶ VEGF indirectly promotes immunosuppression via an abnormal tumour vasculature and directly stimulates a myriad of immunosuppressive cells such as regulatory T cells (Tregs), myeloid-derived suppressor cells (MDSCs) and pro-tumour tumour-associated macrophages (TAMs), while inhibiting antigen-presenting cells and cytotoxic T lymphocytes.⁷ Although VEGF blockade comes across as a favourable anti-cancer mechanism, monotherapy with VEGF inhibitors in HCC has not been shown to be efficacious in achieving optimal outcomes. High doses of VEGF inhibitors excessively prune the vasculature, causes reduction in blood perfusion and thus ensuing hypoxia. Hypoxia stimulates the invasion and metastasis of cancer cells and further drives tumour progression.⁸ On the other hand, immunotherapy in the form of immune checkpoint blockers (ICBs) has shown efficacy in HCC either as monotherapy or in combination with targeted therapies or other ICBs (nivolumab, pembrolizumab plus lenvatinib, atezolizumab plus bevacizumab, nivolumab plus ipilimumab). Best objective response rates to ICB monotherapy ranged from 15 to 23% and increased to about 27% with combination treatment.⁸ A combination of anti-VEGF and immune-checkpoint inhibitor (ICI) appears to a favourable TME to maximally harness the potential of the immune system to eliminate cancer. In this combination, anti-VEGF therapy reverses the immune-inhibitory effect and enables recognition of tumour antigen, TME reprogramming and T-cell recruitment, while ICI restores the T-cell-mediated cancer cell killing. This synergistic immune–vascular interaction has been posited as the mechanism underlying the success of IMbrave150 (atezolizumab plus bevacizumab) trial.⁴ However, more than two out of three of the patients failed to respond in the IMBrave150 and there are no validated biomarkers for atezolizumab and bevacizumab. Our recent work revealed the previously unreported role of VEGF–NOTCH in maintaining an immunosuppressive “onco-foetal ecosystem” in HCC, as evidenced by the co-localisation and interactions among the foetal-associated PLVAP⁺/VEGFR2⁺ endothelial cells (EC), embryonic-like FOLR2⁺/CD163⁺ TAMs and immunosuppressive regulatory T cells such as CTLA4 in HCC (Fig. 1).⁵ Importantly, these results provide additional insights on the mechanisms targeted by atezolizumab and bevacizumab, and highlights the potential value of using onco-foetal biomarkers to stratify patients. In addition, each of these components also represents a potential target to reprogramme the immunosuppressive TME and suggest the potential value of using onco-foetal biomarkers to stratify patients. The understanding of these previously undiscovered mechanisms in HCC can significantly guide the choice of therapies and their combination for potential synergism.

Fig. 1. Onco-foetal ecosystem in HCC.

VEGF–NOTCH maintains an immunosuppressive microenvironment in HCC which is mediated by the co-localisation and interactions among the foetal-associated PLVAP⁺ endothelial cells (EC), embryonic-like FOLR2⁺ tumour-associated macrophages (TAM) and immunosuppressive regulatory T cells.

One of the hallmarks of cancer is the recapitulation of early developmental features in tumour where tumour cells can reprogramme and/or transdifferentiate into cancer stem-like cells.⁹ Similar to embryonic stem cells, cancer stem-like cells have the ability to self-renew and differentiate into multiple lineages. In addition, cancer stem-like cells have also been described to actively remodel the microenvironment, giving rise to a dynamic and heterogeneous tumour ecology.¹⁰ Some malignant cells are known to express antigens specific to foetal development, a classic example being the expression of α fetoprotein (AFP) in HCC, which leads to the use of AFP as the diagnostic biomarker of choice in the management of HCC. The discovery of onco-foetal reprogramming in HCC tumour microenvironment provides critical insights into the previously unexplained link between shared ecosystem between tumour and foetal development and provides unprecedented opportunities for novel therapeutic approaches. In HCC xenograft models, trans-arterial chemoembolisation of recombinant monoclonal anti-PLVAP Fab-T has been observed to induce tumour vascular thrombosis and extensive tumour necrosis, suggesting PLVAP as a viable HCC therapeutic target.¹¹ CD163, which is a marker of inflammatory macrophages, has also been suggested as a potential target for immunotherapy.¹² In the context of the onco-foetal reprogramming as an immune-escape mechanism in HCC, the inhibition of VEGF induced reprogramming of foetal-like endothelial cells and CTLA4-mediated blocking of immunosuppressive phenotype of foetal-like macrophage may confer therapeutic advantages. In addition, the clinical efficacy of CTLA4 checkpoint inhibitors such as ipilimumab has currently been established in advanced HCC when used in combination with nivolumab (CheckMate 040).¹³ Our data points to the involvement of VEGF and CTLA4 in onco-foetal reprogramming mediated immunosuppression of HCC and further supports the use of ICIs in the treatment of HCC.

Taken together, the discovery of the immune-suppressive onco-foetal reprogramming of the tumour ecosystem via VEGF–NOTCH signalling in HCC is potentially valuable for the early diagnosis of HCC and can guide therapeutic strategies. Efforts to develop assays to identify endothelial biomarkers such as PLVAP through non-invasive liquid biopsy may be useful for early diagnosis of HCC and treatment stratification. These discoveries achieved through the comprehensive methodology of multi-region sampling, single-cell sequencing and spatial transcriptomics could also be similarly applied in other cancers.

Author contributions

S.C.C. and S.Y.C. discussed and prepared the commentary. P.K.C. discussed, reviewed and edited the commentary.

Ethics approval and consent to participate

Not applicable.

Consent to publish

Not applicable.

Data availability

Not applicable.

Competing interests

The authors declare no competing interests.

Funding information

Not applicable.