Preclinical and clinical studies show that local, intra-tumoural injection of immunomodulatory agents can result in the shrinking of tumours distant from the site of inoculation. Furthermore, when chemotherapy drugs are used as the immunomodulator, it may circumnavigate the need for systemic chemotherapy delivery. These are the conclusions presented by Thierry de Baere (Institut Gustave Roussy Université Paris-Saclay, Villejuif, France) at the 2020 Symposium on Clinical Interventional Oncology (CIO; every Tuesday of October 2020, online). “Because this is a new field, there are a lot of open questions,” he said. “What is the best target, how should I access the target, is it safe, is it the correct organ, how should I deliver [the drugs], what dose, which regimen, can I monitor the delivery of the drugs?”
“This is a very new, exciting field, and there is a space for interventional radiologists,” he told viewers. Showing a graph depicting the number of new intratumoural immunotherapy trials conducted each year since 1992, de Baere demonstrated the rapid growth of research in this space: from a low of one trial in 1992, to a high of 40 trials in 2016.
When injecting immunostimulatory agents into the tumour, interventionalists hope to achieve local priming, de Baere explained. “We want to trigger a tumour-specific immune response, but we also hope that this [local] response will result in distant effects,” he continued. The idea underpinning human intra-tumoural injections in cancer care is that these “distant effects”—known as abscopal effects—will confer systematic anti-tumour immunity at non-injected tumour sites.
Many different compounds can be used as immunostimulatory agents, and can be used at different stages of tumour immunity. De Baere expanded: “We use products that will release tumour antigens—could be peptides, could be viruses. We want to activate antigen presentation, we want to stimulate cytotoxic cells on site, and we want to deplete or block regulatory T cells [Tregs].”
Providing a more granular breakdown of the agents currently under investigation for drug delivery, de Baere detailed potential combinations for in situ priming of anti-tumour immunity. At every stage of the immune response to cancer, different products are available. For the recruitment of antigen presenting cells, the stimulation of phagocytosis, and the triggering of tumour antigen presentation, interventionalists could consider using toll-like receptor agonists, STING (stimulator of interferon genes) agonists, RIG-I agonists, or anti-C040 agonists. To activate local immunogenic cancer cell death, there are a variety of options, including: the injection of oncolytic viruses, radiotherapy, transarterial chemoembolization (TACE), radiofrequency ablation (RFA), and cryoablation. Anti-PD-I or Anti-PD-LI could be deployed to activate cytotoxic cells, and anti-CTLA-4 could be used to deplete Treg numbers.
Addressing CIO attendees, de Baere said: “Of note, in most of the studies, you are injecting the drug locally, but there is also some intravenous [IV] drugs. Most of the time, you have a PD-LI inhibitor as an IV-combo with drug delivery.”
After detailing the procedural details of intra-tumoural injection, which involve many technical decisions, de Baere stated: “Of course we should be part of this journey, because I think nobody can deliver better local treatment than interventional radiologists.”
A systemic response from a local treatment
De Baere and other interested interventionalists hoping for confirmation of abscopal effects from local intra-tumoural injection are eagerly awaiting the results of ILLUMINATE 301, a randomised, phase III study of tilsotolimod in combination with ipilimumab compared with ipilimumab alone in patients with advanced melanoma following progression on or after anti-PD-1 therapy. Tilsotolimod is a toll-like receptor agonist with potent immunostimulating activity, as described in the literature. In a presentation given by Adi Diab (The University of Texas MD Anderson Cancer Center, Houston, USA) at the European Society of Molecular Oncology (ESMO) annual meeting back in October 2018, the intra-tumoural injection of tilsotolimod combined with ipilimumab resulted in a decrease in the size of an off-target melanoma tumour—one that was not injected with any drugs. Committing this finding to the collective medical memory, Diab and colleagues describe in Annals of Oncology how, in the phase I/II ILLUMINATE 301 study, intratumoural tilsotolimod with ipilimumab was “well-tolerated, demonstrating durable responses (including complete response >21 months), dendritic cell activation, type I interferon response, CD8+ T-cell proliferation in responders, and an abscopal effect”. At the time of writing, phase III results from ILLUMINATE 301 are anticipated in the first fiscal quarter of 2021.
Offering a personal example, de Baere showed the CIO audience a case performed at his own institution. The patient had a single injection of ipilimumab into a liver metastasis, as well as systemic treatment with nivolumab. “You can see we had an overall response from all of the tumours, despite [the fact that] only one tumour was injected.
Intra-tumoural injection potentially most suitable in neoadjuvant disease
Looking at the full spectrum of cancer development, from neo-adjuvant to metastatic disease, de Baere related how he believed that localised, intra-tumoural injections of chemotherapy drugs may be most successful in neo-adjuvant tumours. The interventionalist would inject immunomodulatory agents into a specific tumour, wait a few days or a few weeks to allow for the patient’s immune system to fight the cancer, and then perform surgery. “The reason for that is to prevent distant, metastatic disease to improve the overall survival,” he explained, “and maybe to improve the rate of R0 resection”. R0 resection indicates a cure or complete remission.
Data exist that supports this treatment approach. Presented at ESMO 2019, results of a phase 2, multicentre, randomised, open-label trial of efficacy and safety for neoadjuvant treatment plus surgery versus surgery alone in patients with resectable melanoma found that the combined treatment had improved overall survival and resection-free survival at two years (88.9% vs. 77.4%, p=0.050 and 50.5% vs. 30.2%, p=0.038, respectively). Results from the same dataset presented at the American Society of Clinical Oncology (ASCO) 2019 meeting also report an increased rate of R0 surgical resections.
As this is a new field of study, de Baere stressed the importance of collecting data and following up patients treated via an intratumoural injection of immunotherapeutic agents. “It is important that we use a specific way of looking at the tumour response,” he said. “We need to create a waterfall plots both of the tumours that have been injected, and a waterfall plot of the tumours that have not been injected.” Referencing a Journal of Clinical Oncology (JCO) paper from May 2020, de Baere told viewers that they should follow the intratumoural immunotherapy Response Evaluation Criteria in Solid Tumours (itRECIST) protocol when capturing data from their patients in order to assess any local and systemic responses in a standardised fashion.