NEW YORK – Oncolytics Biotech's virus-based cancer therapy pelareorep (Reolysin) produced positive response rates and preliminary anti-tumor activity in KRAS-mutated metastatic colorectal cancer patients, according to data from a Phase I trial published in March in Molecular Cancer Therapy.
This trial is part of the company's larger development program to advance pelareorep in several biomarker-defined patient populations, including a lead program in breast cancer. In fact, within Canada-based Oncolytics there is a strong focus on the discovery and validation of biomarkers that can improve patient outcomes, help the company run more efficient clinical trials, and facilitate risk-sharing contracts with payors.
For example, in the recently published Phase I dose escalation trial, researchers focused on evaluating the efficacy and safety of colorectal cancer patients with KRAS mutations after earlier pre-clinical models showed the therapy was selectively toxic for KRAS-mutant cells over KRAS wildtype cells, and its effects were enhanced with FOLFIRI. Investigators in the trial microscopically confirmed that these viruses homed in on the tumor cells.
The Phase I study enrolled 36 metastatic colorectal cancer patients with a confirmed KRAS mutation who progressed on prior oxaliplatin-based chemotherapy. During the trial, patients received chemotherapy regimen FOLFIRI, VEGF-inhibitor bevacizumab, and pelareorep. Participants were divided into five cohorts of three to six patients, each receiving a different dose to determine a maximum tolerated dose and dose-limiting toxicities with the combination of the drugs and the optimal dose for Phase II.
Researchers led by Sanjay Goel at Montefiore Medical Center reported that out of 30 evaluable patients, six patients had a partial response to treatment and 22 had stable disease. The median progression-free survival was 22.2 weeks and the overall survival was 11.7 months.
In the cohort that received the recommended Phase II dose, three out of six patients had a partial response. These patients had a median progression-free survival of 65.6 weeks and an overall survival of 25.1 months.
Pelareorep is derived from a reovirus whose genetic elements are made up of double-stranded RNA. Double-stranded RNA is something that's recognized by normal cells as something foreign that should be eliminated. When a reovirus is put into a healthy individual, it will get picked up into various cells where it will try to replicate and release its double-stranded RNA into the cell. This in turn will trigger proteins in the cell that act like a home surveillance system, constantly looking for foreign entities. When a protein called PKR, or dsRNA-activated protein kinase, sees double-sided RNA, it shuts down translation in that cell so that the virus can't make copies of itself.
"This is why reovirus doesn't really pose a risk to us," Matt Coffey, CEO and co-founder of Oncolytics, explained. This is reflected in the Phase I colorectal trial, where favorable safety profiles were observed for patients. "The story becomes a little bit different when we have cancer tissue."
Cancer cells lose a lot of functions along with some of the cell's normal housekeeping genes. In cells with oncogenic mutations, such as KRAS and others that could cause aberrations in the RAS signaling pathways, the activity of the PKR protein gets disrupted. When a reovirus is put into cancer cells, the viral replication process is unchecked, and results in an accumulation of the virus along with its double-stranded RNAs, which creates an alarm signal in the cell. The cell reacts to this by releasing chemical signals called chemokines, and cytokines, and interferons, which in turn triggers cytotoxic T cell responses.
"What the virus is able to do is selectively infect tumor tissue. It's the infection process that alerts the immune system that the tumor is a foreign entity," said Coffey. At first, the team thought pelareorep worked by lysing the cells, but they now think the virus is labeling the cancer cells for the immune system to find. Coffey also offers the analogy of bank robbery, where the cancer cells are the anonymous robbers that have been able to evade security and slip into the bank, and pelareorep is the dye pack that's taken up without the robber being aware. But once it explodes, it can alert security to who the robber is.
"All of a sudden, this person who before looked like everybody else, now looks very different," Coffey said. "The virus can initiate a very similar process through your immune system. It takes a tumor, which was first anonymous, and makes it obvious to the immune system … that it shouldn't be there."
The Molecular Cancer Therapy study also observed the activation of natural killer cells, akin to the nonspecific beat cops of the immune system, which come rushing towards the chemical signal emitted by these cancer cells and destroy the tissue that's producing it. Then, researchers observed the kicking in of the adaptive immune response, which is a learned response where more specialized immune cells come and survey the debris composed of tumor tissue and viral fragments. In this way, the virus almost acts like a vaccine against the cancer, said Coffey.
"We get tumor-infiltrating lymphocytes rushing into the tumor and they actually learn what the tumor looks like. And what we're finding in patients who've been successfully vaccinated is that we can detect these memory T cells that now recognize the tumor and hopefully prevent recurrence," he said. "We only saw that [activation of the innate and adaptive response] on the cycles where we give virus, but not on the cycles where we only gave chemo. So, we know it's the virus that's doing this."
Advancements in breast cancer
The team at Oncolytics has observed this effect not only in the KRAS-mutated colorectal cancer cohort, but also in their lead program for pelareorep in hormone receptor-positive, HER2-negative, metastatic breast cancer patients.
Data from a randomized Phase II study of patients with metastatic breast cancer published in 2017 indicated that women who had hormone receptor-positive and HER2-negative breast cancer had a doubling of overall survival in the test arm with pelareorep and chemotherapy over the control arm with chemotherapy alone.
While response rates and progression-free survival didn't significantly differ between the two arms, in the control arm, women had about a 10-month survival, but in the pelareorep arm, that got bumped up to nearly 21 months.
Oncolytics is now gearing up for a Phase III study of the viral-based therapy in HR-positive, HER2-negative breast cancer. In preparation, they are working with the US Food and Drug Administration on the trial design and have partnered with Roche, Pfizer, and the cooperative group SOLTI on two additional clinical trials that will help identify whether to add a checkpoint inhibitor to the Phase III study. In this Phase III study, researchers will also validate the biomarkers of response.
The study with Roche and SOLTI is designed to gain further insights into how pelareorep works to engage the immune system in metastatic breast cancer. The trial is currently enrolling in Spain.
"We're seeing a bit of a delay in enrollment because of the COVID outbreak. But, fortunately, the numbers in Spain look like they're leveling off, so we're anxious to get back to work," said Coffey.
The first part of the collaboration study is aimed at validating a biomarker based on T cell populations to predict which breast cancer patients are likely to respond to pelareorep within three weeks of treatment. The second part of the study will gauge whether pelareorep's activity could be enhanced with the addition of the PD-L1 checkpoint inhibitor atezolizumab (Roche's Tecentriq).
Coffey said that Oncolytics is planning to release data from this study in May at the European Society for Medical Oncology (ESMO) Breast conference. But they are already starting to see some preliminary evidence suggesting biomarker efficacy.
"What we're finding is women who have been successfully vaccinated with the virus will develop brand new T-cells, hundreds of T-cells that can recognize and kill the tumor," said Coffey. That change can be detected through patient blood samples via T cell receptor (TCR) sequencing performed with Adaptive Biotechnologies' immunoSEQ Assay. Each T cell clone is unique, and the assay can identify all the T cell clones a patient harbors at baseline.
"At baseline, patients who are unlikely to respond have unfortunately very little reserve of immune cells left, and what you get is low levels of various clones and no predominant clones that would suggest that you're having an immune response against your tumor," Coffey explained. "In the patients that respond very favorably, what we see is that at baseline, they have a lot of variability or higher clonality, which means that they actually have a nimble immune system that is still functioning at a high level and can actually respond to threats."
The TCR sequencing assay can be taken at baseline but can also be taken again at three weeks post-treatment. In the patients who've been successfully vaccinated, Coffey said that they would expect to see the creation of hundreds of brand new T cell clones, which can be further quantified. And if that number falls above a certain threshold, researchers can say the patient had a good vaccination effect. If that number is below the threshold, it means the patient was unlikely to have responded to the immune agent. The team is currently validating this marker in breast cancer and working on determining an optimal cutoff.
Oncolytics is in discussions with Adaptive Biotechnologies about potentially making the technology into a companion diagnostic for pelareorep.
A test that provides early insights into whether a patient will likely respond to a very expensive treatment can help Oncolytics ink risk-sharing reimbursement arrangements with payors. "From a pricing perspective, a course of treatment [with checkpoint inhibitor CTLA4] takes over $100,000-$150,000. Adding another immunotherapy therapy to that could be quite expensive," said Coffey. "But if you know the patient's having a response and doing well early, it's easy to get the reimbursements."
Coffey noted that a lot of immunotherapy interventions often don't cause massive changes in objective response. But, for good readout about overall survival, some patients can be on studies for months and not be deriving any benefit. This can cut into the time they have left, during which they could be getting other therapies that work better.
"When we move on to more personalized medicine, as it becomes more expensive, it becomes absolutely critical to identify who's going to respond and then confirm response very early," said Coffey. "One of the possible strategies we discussed is with an assay like the one we have in hand now, we could even [give patients] a cycle [of treatment] for free to see whether or not the person will have a response, and then charge the patients who are benefiting and get the other patients off and onto something more effective for them."
Similar objectives are being studied in the second trial Oncolytics is conducting with Pfizer. Patients with HR-positive, HER2-negative breast cancer, will be sorted into three arms: a control arm with only the chemotherapy paclitaxel, a treatment arm with paclitaxel plus virus, and an additional arm with paclitaxel, virus, and PD-L1 checkpoint inhibitor avelumab (Pfizer and Merck KGaA's Bavencio). Investigators will also look to confirm the biomarker with TCR sequencing.
Oncolytics has recently started activating sites for the study and anticipates dosing the first patient in Q2.
Bolstering response to checkpoint inhibitors
Of course, Oncolytics is not the first company to realize potential in combining checkpoint inhibitors with cancer-targeting viruses. Viruses have in the past been shown to be adept at helping cancer patients respond better to immune checkpoint inhibitors.
Amgen is currently conducting a Phase III trial with herpes-virus derived treatment T-Vec (Amgen's Imlygic), to determine how it works in metastatic melanoma in combination with the anti-PD1 checkpoint inhibitor pembrolizumab (Merck's Keytruda) versus pembrolizumab alone.
Checkpoint inhibitors that target PD-1 or PD-L1 don't work on the tumor, they work on the immune system. If PD-1 or PD-L1 is expressed on tumor cells or on T cells, then these agents work incredibly well, said Coffey. But for them to work, the tumor has to express PD-L1. "If it doesn't, there's nothing for these antibodies to work on," he said. Pelareorep also generates autoreactive T cells, which checkpoint inhibitors require to work.
"The virus will create these new T-cell clones. So, even if the patient doesn't have them at baseline, we can create them so that the checkpoint inhibitors can work better," Coffey said. On top of that, the tumor cell infection process releases a chemical called interferon. Research has found that interferon can cause the over-expression of PD-L1 on tumor targets. And because the virus has caused an actual infection in the body, it's able stimulate a pro-inflammatory response that drives all these T cells into the tumor tissue.
"We think we can hopefully cause a 30 percent to 50 percent increase in the number of patients that can respond to these [checkpoint inhibitor] therapies by basically just pre-treating with the virus," said Coffey. "We're exploring that in those two breast cancer studies before we initiate the Phase III, which we anticipate doing next year."
The Phase III trial would be the registration-enabling program in breast cancer. It will likely start enrollment in 2021 or early 2022.
Designing successful studies
Coffey projected that a registration program for pelareorep in GI malignancies, which the firm will investigate in an upcoming basket trial, will come six to nine months after the breast cancer program.
For GI malignancies, such as colorectal and pancreatic cancer, Coffey mentioned that they are working with industry and cooperative groups to design a basket study with the same goals of carving out responding populations through biomarker identification.
"What's nice about those basket studies is when you identify a signal, you can very quickly push them into a Phase II/III registration pathway," said Coffey.
At the American Society of Clinical Oncology GI meeting this year, researchers nominated a new predictive biomarker for pelareorep sensitivity. In a pancreatic cancer study that was run by the National Cancer Institute, researchers found that super responders to the pelareorep and chemotherapy treatment all expressed a protein called CEACAM6 at a low level.
CEACAM6 is an immunomodulatory carcino-embryonic antigen-related cell adhesion molecule that plays a role in viral trafficking. Patients on the pelareorep arm who had low levels of CEACAM6 saw a doubling in progression-free survival compared to patients on the control arm, where the CEACAM6 level wasn't associated with their treatment outcomes.
Still, across the different cancer indications that pelareorep has been tested in, Coffey has noted some unmistakable universality to the TCR-defined biomarker. At a poster presented at the annual American Association for Cancer Research meeting in 2019, the team found that the TCR biomarker they are developing in breast cancer can also be used to predict response in pancreatic cancer.
On the whole, Coffey said that the TCR sequencing-based biomarker has allowed the company to design and run more successful studies, by enriching for patients who are likely to do well on treatment and sparing those who are not. "We're just designing a protocol now that would allow us to look at various indications in the context of a checkpoint inhibitor," said Coffey. "[The TCR sequencing] allowed us to quickly identify where we're seeing activity in GI cancers and move that into Phase II/III development, or quickly get rid of indications where it doesn't look like it's going to be successful."