Combination of immuno-oncology and liver-directed therapy: Where do we stand?

Bridging the gap between the IOs: “We need to determine whether to freeze, fry, shake, shock, zap, or choke a tumour,” writes Brad Wood

Ushering in “the era of IO”, Brad Wood calls for interventional oncologists to learn the language of immuno-oncology—“the other IO”—in order to improve care for cancer patients. Discussing the “speculative but huge potential impact” of combining interventional radiology (IR) approaches with an understanding of immune-oncology, Wood warns readers that this is not merely an exciting opportunity to increase the scope of interventional oncologists’ work, but a necessary step in ensuring the discipline stays at the vanguard of oncologic treatment. immuno-oncology

It has been known for decades that local-regional therapies like ablation and embolization can stimulate a tumour-specific systemic immune response, or even, on some occasions, tumour regression remote from the treated area. What is not known is who, what, where, when, and how to make it happen more often. Enter the era of IO, which to most people means immuno-oncology (IO), and not interventional oncology (IO). Bridging the gap (between IO and the other IO) requires learning the language and themes of cancer immunology. Interventional radiology (IR) has several hammers to pull out of our big IO toolbox. However, we need to determine whether to freeze, fry, shake, shock, zap, or choke a tumour—that is, whether to use cryoablation, radiofrequency ablation (RFA), microwave ablation (MWA), irreversible electroporation (IRE), high-intensity focused ultrasound (HIFU), histotripsy, transarterial chemoembolization [TACE], or Yttrium-90 (Y-90) radioembolization. We also need to determine which histology to treat, at what stage, alongside what IO drug, and in what sequence. Where in the cancer immunity cycle do the IO tools immuno-modulate? How do we balance pro-immune versus pro-metastatic pathways? The window of opportunity remains open for answering the many questions in a rational way.

Checkpoint inhibition takes the brakes off our natural immune response to tumours. Although revolutionary and Nobel-prize worthy, checkpoint inhibitors usually fail as monotherapy, are costly, and are not without side effects. Enter IR/IO. Local IO therapies alongside of checkpoint inhibition can convert an “immune-cold” tumour into an “immune-hot” tumour, potentially turning local into systemic. Immunomodulatory mechanisms boil down to enhancement of: 1) antigen release and presentation, 2) antigen-presenting cell (APC) maturation, 3) T-cell priming proliferation and trafficking, and, 4) reduction of immune tolerance and immune resistance. The molecular tumour micro-environment and the dynamic immune compartments (blood vessels, lymph nodes, target tumour, and remote tumours) also merit consideration.

Initial clinical experience with checkpoint inhibition drugs plus ablation or embolization for advanced hepatocellular carcinoma (HCC) was first reported over five years ago in a National Institute of Health (NIH) study.1, 2 This showed disease stabilisation in the bulk of patients with advanced and heavily pre-treated HCC via this combination approach, often with intentional subtotal treatment of the target tumours. This study also has strong blood and tissue evidence for an immune mechanism when looking at responders versus non-responders. For cholangiocarcinoma, a similar approach with subtotal ablation can be taken.3 Human tissue and survival data convince more medical oncologists at a multidisciplinary tumour board meeting than curing mice in 1,000 lab studies. In this setting, the local or regional therapy is meant to boost the immune system, deposit antigens, promote APC and T-cell maturation, broaden the T-cell receptor response, and tip the balance of immune-regulation away from immune tolerance and towards tumour-specific T-cell activation. More recently, Bevacizumab plus Atezolizumab was approved by the US Food and Drug Administration (FDA) for locally-advanced HCC,4 and ongoing clinical trials will hopefully address whether and when local ablation or embolization add value. The therapeutic options are evolving fast, so IR needs to organise a clinical trial infrastructure, so we can ask hypothesis-driven questions in a standardised fashion. Team science wins.

Brad Wood

Emerging data support aggressive local-regional therapies when immunotherapy fails locally in established tumors (versus new widespread metastases). Specifically, when stratified by patterns of failure, local therapy has been shown to be particularly effective after failed checkpoint inhibition, supporting an aggressive re-treatment approach, when progression is confined to established tumours (rather than new breakthrough metastases).5 Like ablation, drug-eluting bead TACE (DEB-TACE) and Y-90 radioembolization reduce regulatory T-cells (good) and expand T-cell responses in HCC. Doxorubicin potentiates immunogenic cell death and enhances cross-priming, but ischaemia cascades turn on damage/pathogen-associated molecular patterns (DAMPS/PAMPS), HIF1a, and cell stress pathways. IRE may leave vessel highways intact for APC and T-cell trafficking, which may explain why IRE (shock) may be better for immunomodulation than RFA (heat) or cryoablation (freeze) for T-cell activation. Histotripsy may be better than heat when combined with checkpoint inhibition.6 Both IRE and cryoablation release more proteins than heat. Yet RFA heat may have a favourable impact upon immunosuppressive T-regs than cryoablation. Cryoablation may have the most inflammation and cell stress signals. Any ablation might recruit more good macrophages than bad, when the margin interacts with the tumour (thus leaves tumour behind). Yet colorectal metastases may explode when partially treated and exposed to the hypoxia of partial treatment.

Clearly, we need to biopsy more after IO therapies. We need to learn more about the tools to measure immune effects (such as immunohistochemistry, flow cytometry, and sequencing). It is a great time to have coffee with an expert from another discipline!

So when you freeze, fry, shake, shock, zap, or choke the tumor—treat the patient, not the picture. Do not be a technical robot and just treat geometric spots in the liver. Think before ablating. Create the proper tumour micro-environment via timing, sequence, and delivery of rational IO plus IO immunomodulation. This is a one-time opportunity, with speculative but huge potential impact. New immuno-oncology trials with checkpoint inhibitor drug combinations have exploded in the past two years.7 Where is IR? We need to be at the table and not on the menu.

Brad Wood is an interventional oncologist at the National Institutes of Health (NIH) Center for Interventional Oncology, Bethesda, USA.


  1. Presentation at the American Society of Clinical Oncology (ASCO) Annual Meeting, 2015
  2. Duffy AG et al. Tremelimumab in combination with ablation in patients with advanced hepatocellular carcinoma. Journal of Hepatology. 2017 Mar;66(3):545–551. Doi: 10.1016/j.jhep.2016.10.029
  3. Changqing Xie et al. Tremelimumab in combination with microwave ablation in patients with refractory biliary tract cancer. Hepatology. 2019 May;69(5):2048–2060. Doi: 10.1002/hep.30482
  4. Richard Finn et al. Atezolizumab plus Bevacizumab in unresectable hepatocellular carcinoma. New England Journal of Medicine. 2020 May; 382:1894–1905. Doi: 10.1056/NEJMoa1915745
  5. Nicholas Klemen et al. Patterns of failure after immunotherapy with checkpoint inhibitors predict durable progression-free survival after local therapy for metastatic melanoma. Journal of Immunotherapy of Cancer. 2019 July;7:196. Doi: 10.1186/s40425-019-0672-3
  6. Eranki A, et al: High-Intensity Focused Ultrasound (HIFU) Triggers Immune Sensitization of Refractory Murine Neuroblastoma to Checkpoint Inhibitor Therapy. Clinical Cancer Research 2020 Mar 1;26(5):1152-1161. doi: 10.1158/1078-0432.CCR-19-1604
  7. Greten TF, et al: Combined locoregional-immunotherapy for liver cancer. Journal of Hepatology 2019 May;70(5):999-1007. doi: 10.1016/j.jhep.2019.01.027


Please enter your comment!
Please enter your name here