NEW YORK – Immunotherapy is the most exciting advance in cancer treatment in recent years but not everyone responds to these drugs. As researchers explore different ways of improving patients' ability to respond to these new therapies, some are turning to an idea that is hundreds of years old: viruses.
Immune checkpoint inhibitors work by amplifying or unleashing cancer patients' ability to respond by taking the brakes off their immune system and letting it fight the cancer. But the magnitude of that response depends on the presence of antigens in the body for the immune system to recognize. Researchers have been working to identify biomarkers that can predict which patients are most likely to respond to immunotherapies, but the biomarkers identified to date, like tumor mutational burden, aren't useful in all cancer populations.
Pediatric cancers, for example, haven't been helped much by checkpoint blockade therapy because unlike adult tumors, they tend to lack a large number of tumor mutations that their immune systems can recognize as antigens.
And it's not only pediatric cancer patients who experience this problem. Any immunogenically "cold" tumor — those that have fewer mutations or lack substantial T cell infiltration — such as gliomas, would likely exhibit the same non-response to immune checkpoint blockade therapy.
In addition to predicting which patients will and won't respond to immunotherapy, researchers are also working on strategies to bolster the immune responses of non-responders.
The idea of using viruses to treat cancer is "ancient, but only now is it getting serious," said Matthias Gromeier of the Duke Cancer Institute. The creation of immune checkpoint blockade therapies and their increasingly widespread use has opened the door to the oncolytic virus field, he added. In the last several years, oncologists have learned more about the immune system's interaction with cancer, and are exploring novel ways to overcome the immunosuppressive and anti-immune mechanisms that cancers use to survive.
As of now, there's only one oncolytic virus treatment approved by the US Food and Drug Administration for the treatment of cancer. In 2015, the agency approved talimogene laherparepvec (Amgen's Imlygic), known as T-Vec, for the treatment of unresectable melanoma. T-Vec is a genetically engineered herpes virus combined with a modified version of the human gene that produces the GM-CSF glycoprotein. The drug, which is meant to be injected directly into the cancerous lesions, works by replicating in cancer cells and causing them to burst.
Some researchers are looking for ways to improve on this strategy by using biomarkers to design oncolytic virus therapies that would target certain biological mechanisms driving tumors. For example, scientists at Okayama University in Japan published a study in Cancer Research in July describing an oncolytic virotherapy they're developing to treat pancreatic ductal adenocarcinoma by eliminating p53-inactivated malignant tumor cells.
Another promising avenue of research involves combining oncolytic viruses with immunotherapy. "[Immune checkpoint blockade treatments] really have provided a fertile ground to explore immunotherapy in completely new ways," Gromeier said. "Viruses fit very logically with immune checkpoint inhibitors."
The research is still in early stages, and it is unclear whether oncolytic viruses will work as a treatment in all cancers, or if the presence or absence of viral receptors in certain cancer cells will require a more precise approach. The potential for adverse events will be another important consideration when deciding which patients to treat.
"We probably will be able to predict who will respond [to oncolytic virus therapies]. It's not simple, but there are good indications that we can [use to] test the ability of a patient's tumor to respond to immune stimulus in general," Gromeier said. "There are immune markers that probably we will be able to [use], at some point in the future, to tell if patients are responding or not."
Before they develop biomarkers for response, however, researchers must find the best oncolytic virus treatment strategy for each type of cancer. "There are several other oncolytic viruses in various stages of clinical development," Gromeier said, highlighting that his group is conducting clinical trials using a recombinant polio virus in brain tumor patients. Melanoma is a particularly fertile ground for investigating oncolytic viruses because it is relatively easy to inject the viruses directly into the tumors, where they have been shown to work most effectively, avoiding systemic administration.
Across these different investigations, researchers are interested in the same question: can viruses make a patient's immune system attack the tumor?
"The objective with most of these oncolytic virus trials is to arouse the immune system to the presence of the tumor, meaning that you have an infection of the tumor that hopefully will lead to the host immune system rejecting that tumor or fighting that tumor off," Gromeier explained.
New research has shown that the administration of these oncolytic viruses into tumors that are immunogenically "cold" can make them "hot," increasing a patient's response to immune checkpoint inhibition. That's the idea behind the work of Gustave Roussy researcher Aurélien Marabelle, who presented his investigation into the combination of rotavirus with immune checkpoint blockade at the annual meeting of the European Society of Medical Oncology in September and in a study published recently in Science Translational Medicine.
In his presentation, Marabelle said childhood cancers are a model of tumors that are immunogenically cold, but that inflaming these cancers could lead to a response to immune checkpoint blockade in pediatric patients. He chose to work with rotavirus in mouse models of neuroblastoma, a cancer that is "extremely cold" and shows no response to immunotherapy alone or in combinations.
"We had the idea of using anti-infectious vaccines as a source of pathogens, of pro-inflammatory products," Marabelle said in an interview. The researchers are screening a variety of FDA-approved vaccines for their ability to stimulate toll-like receptors, a class of proteins that play a key role in the innate immune system. They chose to work with the rotavirus vaccine because it was triggering NF-κB signaling downstream of the toll-like receptors, but without the necessity of having toll-like receptors expressed on the cell.
NF-κB is a protein complex that controls transcription of DNA, cytokine production, and cell survival, and is involved in cellular responses to stimuli such as bacterial or viral antigens. It also plays a key role in regulating the immune response to infection, and incorrect regulation of NF-κB has been linked to cancer, among other diseases.
The researchers also observed that a high concentration of the vaccine was killing cancer cells in the cell lines. So, they obtained pure rotavirus, and tested it on both pediatric and adult cancer cell lines. "We could show that they had oncolytic properties; that they could, in fact, destroy these cancer cells and also induce immunogenic cell death," Marabelle said.
The team then moved on to test rotavirus in murine models of neuroblastoma, which is refractory for immune checkpoint targeted antibodies. In some mice, simply the injection of rotavirus into the tumor was enough to cure the mice. But when they combined the intratumoral administration of the rotavirus with anti-PD1, anti-PD-L1, or anti-CTLA4 therapies, the researchers found that they could cure all the mice in their model, regardless of the specific immunotherapy they chose to use.
"We had a clear synergy with the combination of intratumoral rotavirus plus systemic immune checkpoint antibody," Marabelle said. "Although the immune checkpoint antibody alone would have no activity, the intratumoral rotavirus was able to overcome the resistance to immune checkpoints and also to synergize with them."
The researchers also conducted experiments in mice with two tumors. They found that administration of rotavirus into only one of the tumors was enough to get the mice to respond to immune checkpoint blockade, and that both the injected tumor lesion and the distant tumor lesion shrank. Surprisingly, they further found that this response was repeated even when they used inactivated rotavirus, as well.
When they looked for a biological explanation for the synergy between the rotavirus and the immune checkpoint inhibitors, the researchers found that there was a critical innate immune sensor called RIG-I, which recognizes patterns of receptors for viruses such as influenza A, Sendai virus, and flavivirus in the cytoplasm of cells.
"When these sensors detect the nucleic acids, the RNA from the viruses, then they trigger a pathway that will eventually accumulate type I interferons," a large subgroup of interferon proteins that help regulate the activity of the immune system," Marabelle said. "So, we found that what was critical was actually for the rotavirus to trigger this RIG-I pathway, [stimulate] type I interferons, and then have the synergy with immune checkpoint antibodies — it was really the pro-inflammatory feature of the virus through the RIG-I pathway that was critical to get the synergy with immune checkpoint antibodies.
Indeed, Marabelle added, because the efficacy of the rotavirus therapy was dependent on the RIG-I receptor and type I interferon pathway, mutations in these areas could be used as biomarkers to predict whether patients will respond to the therapy. "Patient selection based on the absence of these molecular alterations would make sense," he said, noting that such biomarkers with a negative predictive value can sometimes work better for immune checkpoint therapy than biomarkers with positive predictive value.
In his own work with polio in brain tumors, some of which was published in the New England Journal of Medicine in July 2018, Duke's Gromeier said that polio virus receptors are expressed on tumor cells themselves and macrophages — immune cells that are present in large numbers in brain tumors — both of which are important to how the cancer responds to immunotherapy.
"With the polio virus work, [we found that] the virus cannot kill off the cancer by itself. It just can't. It might kill some tumor cells, but it cannot eliminate whole big tumors," Gromeier said. "What it can do is infect a certain number of cancer cells and damage these cells. And at the same time, it infects immune cells and it activates these immune cells. And what we have described in animal models and in laboratory research is that when you put these two together, you really get a very powerful immune response against the tumor."
According to the research Gromeier and his colleagues have done, injecting the polio virus directly into brain tumors has caused better results in patients than the standard of care, which consists of surgical resection, radiation, and chemotherapy.
"This therapy almost always fails eventually, and patients' tumors return. And when the tumor returns, there's no more therapy," Gromeier added. "These are horrible cancers — they're very lethal, they're rapidly progressive. And we really have nothing [to treat patients with]. And so those patients are the ones we treated with the virus, and I believe that our clinical data showed [that we could] generate patient immune response against a tumor [using polio]."
Importantly, both Marabelle and Gromeier showed that previous exposure to or vaccination against rotavirus and polio didn't impede the synergy of the viruses with immune checkpoint inhibitors or the reaction of cancer cells to the combination treatment.
This lines up with what was shown during the development of T-Vec, Marabelle added. People with preexisting immunity against the herpes virus still saw benefit from treatment with T-Vec.
There's still a lot of research that needs to be done before oncologists start injecting patients with viruses, however. The most important thing is to start testing in humans as much as possible. Marabelle emphasized that his work was preclinical, and that there is no human clinical data on the use of rotavirus for the treatment of neuroblastoma.
In his presentation at the ESMO meeting, he also said that progress in this field will require learning exactly what level of inflammation in cold tumors would be therapeutic for the patients, and how much inflammation is needed to synergize with immunotherapy.
Marabelle also said oncologists will need to identify the most rational combinations of immunotherapy and oncolytic viruses. Gromeier agreed, saying that the most rational way to apply these viruses in cancer patients is to combine them with immune checkpoint inhibition, rather than treating them as potential monotherapies.
"I think in the end, we will combine them. We will need to. Almost all cancer therapy is some form of combination — we combine radiation and chemotherapy or several kinds of chemotherapy because it's just really hard to find a weapon that will work well enough on its own," he said. "The immune system is very complicated and while viruses can do certain things very well, certain other things they cannot. So, it's a very attractive possibility for oncolytic viruses [to be combined with immunotherapy]. Quite a number of them can be exploited."
Researchers will also have to determine the possibilities for toxicity from the use of oncolytic viruses and determine if they outweigh the benefits. Gromeier noted that his work has shown that polio causes some inflammation in brain tumors. But while this is a beneficial reaction to a certain extent, inflammation in the brain can also cause additional problems for the patient. Further, he added, some patients have been shown to be resistant to the effect of certain viruses if they have a preexisting immunity — further research is needed to determine which viruses this will work against.
And lastly, he said, safety must be paramount. "There have been some quite concerning safety issues with certain viruses in Phase I safety trials," Gromeier added. "This is not some kind of cowboy field of science where we just need to rush ahead. This has to be done very carefully and deliberately, and safety first."
Marabelle and Gromeier are now pursuing further research on rotavirus and polio, respectively. Gromeier's team is conducting a clinical trial of polio in melanoma patients, and an early trial in breast cancer, as well. They believe they can use the polio virus to treat just about any cancer because of the near-universal presence of polio virus receptors in cancer cells.
"That will require some proof. Ultimately, you have to do this in clinical trials — the animal models and the other systems we have are not always good enough to really predict the outcomes," he said, echoing Marabelle. "We will have to potentially perform quite a large number of clinical trials to come to some answers. It's possible that certain viruses have better properties in certain tumors. That all will need to be figured out. We're still very early [in our research]."
In addition to continuing his work with rotavirus, Marabelle is also participating in trials on molecules that can trigger the type-1 interferon pathway, including a trial involving a RIG-I agonist.
And while they work, researchers are also keeping an eye out for results from a Phase III trial Amgen is currently conducting with T-Vec, to determine how it works in metastatic melanoma in combination with the anti-PD1 checkpoint inhibitor pembrolizumab (Merck's Keytruda) versus pembrolizumab alone.
"It will be very important for the future of the oncolytic virus field because either this trial is positive, showing that the combination will be synergistic, more efficacious in patients rather than anti-PD1 alone, and you will see a lot of interest in that kind of approach," Marabelle said. "Or the trial will be negative, and it will have a detrimental impact on the strategy."
Results from the Amgen trial are expected to be published in 2020.