Injecting cancer treatments such as vaccines and genetically modified bacteria and viruses directly into tumors may shrink or destroy the tumors and stimulate an immune response that attacks tumor cells. These treatments have shown promise in clinical trials, and newer trials are exploring whether such treatments can augment the effectiveness of immunotherapy in patients with metastatic or inoperable disease.
“I think we are poised to make a big difference for patients with unresectable or other difficult-to-treat tumors,” said Vivek Subbiah, M.D., an assistant professor in the Department of Investigational Cancer Therapeutics at The University of Texas MD Anderson Cancer Center.
Dendritic cell vaccine
Dr. Subbiah is the principal investigator of a phase I/II clinical trial (2013-0160) of a dendritic cell vaccine in patients with locally advanced or metastatic solid tumors. The personalized vaccine is made from each patient’s dendritic cells and injected into that patient’s tumor.
The phase I portion of the trial found no dose-limiting toxic effects. The main side effect was fever, which was manageable, although one patient experienced systemic inflammatory response syndrome necessitating emergency care.
Thus far, Dr. Subbiah said, analyses of tumors injected with the vaccine found increases in tumor-infiltrating lymphocytes. However, in the current trial, only one tumor per patient is injected, and patients with multiple tumors have had stable disease at best in their noninjected tumors.
“Biopsies of noninjected tumors showed some distant effects, but these effects were weak,” said Ravi Murthy, M.D., a professor in the Department of Interventional Radiology, who performs the image-guided intratumoral injections for patients in the trial. “In the next study, we’ll inject multiple tumors—perhaps up to five sites—multiple times.”
Another strategy for future trials might be to combine an intratumoral vaccine with an inhibitor of an immune checkpoint such as programmed cell death protein 1 (PD-1) or the PD-1 ligand (PD-L1). Dr. Subbiah said that such combination therapy might create both local and systemic effects.
The concept of using bacteria to treat cancer may sound radical, but it is hardly new. In the 1890s, William B. Coley, a surgeon and pioneer of immunotherapy, observed that cancer patients with bacterial infections sometimes experienced tumor regression.
But until recently the risks associated with bacterial infection outweighed the benefits of bacteria-based treatment for cancer patients.
A nontoxigenic strain of bacteria, Clostridium novyi-NT, was engineered to be capable of germinating only in a hypoxic environment such as that in some tumors. “C. novyi-NT is not capable of germinating in normal tissue because the oxygen levels are too high,” said Filip Janku, M.D., Ph.D., an assistant professor in the Department of Investigational Cancer Therapeutics. “But C. novyi-NT is capable of germinating in hypoxic cancer tissue and causing lysis—the destruction of the tumor.”
Hypoxic tumors historically have been difficult to treat. The limited blood supply to the tumors hampers the delivery of systemic agents. Also, the lack of molecular oxygen, which acts as a radiosensitizer, limits the efficacy of radiation therapy against hypoxic tumors. Bacteria that target such tumors could therefore be very useful.
In the first clinical trial of C. novyi-NT, which was not conducted at MD Anderson, the bacteria were injected intravenously. “The theory was that the bacteria would be selectively delivered to the tumor, which did occur; however, the problem was that some tumors were not accessible by surgery, which made adverse events difficult to manage,” Dr. Janku said. “It was felt that intratumoral injection would work better because we can select a lesion that is accessible so that we can control for potential collateral damage.”
C. novyi-NT is injected intratumorally in an ongoing multicenter trial (2013-0549) for which Dr. Janku is MD Anderson’s principal investigator. The C. novyi-NT injection is an outpatient procedure, but the study’s protocol requires patients to be hospitalized for 7 days after the injection so that immediate care can be given in case of adverse events. As a precaution, all patients receive oral doxycycline starting on the seventh day after C. novyi-NT injection.
Although preliminary data from the current trial are not yet available, Dr. Janku estimates that some level of local response is seen in about one-third of patients. Some of these responses have been dramatic, with complete destruction of the tumor in a few days.
Side effects of the treatment can be dramatic as well. For example, the first patient treated in the trial had a tumor surrounding the right humerus. The tumor was injected with C. novyi-NT and was destroyed, but the patient suffered a right humerus fracture 2 months later because the bone was too fragile without support from the large surrounding tumor. “We had a bad consequence of a good effect,” Dr. Janku said. Other patients had systemic inflammatory reactions that included fever, low blood pressure, and coagulopathy; these reactions were controlled by antibiotics and other supportive measures. “The reaction seems to be mediated by cytokines rather than bacteremia because we don’t see bacteremia in these patients,” Dr. Janku said.
Researchers think the bacteria works through two mechanisms. The first, direct lysis of the tumor, is very rapid. The second is the abscopal effect: the immune system is primed by the destruction of the tumor to recognize the cancer. “It’s still too early to tell, but some patients seem to have signals of a systemic response,” Dr. Janku said. “We see a slowdown in the growth of the tumors that were not injected, but that effect doesn’t last very long. Data suggest that this therapy can be even more effective when combined with some immune checkpoint inhibitors.”
C. novyi-NT will be combined with a PD-1 inhibitor in an upcoming trial at MD Anderson and other centers. Patients who have superficial masses or accessible masses that do not involve the bones or internal organs and who are not candidates for surgery may be candidates for the trial. It is hoped that the immune checkpoint inhibitor will enhance the systemic response to the bacteria.
“The key is to have enough bacterial germination for a maximal antitumor effect and at the same time to have a systemic reaction that is controllable and does not put patients at risk,” Dr. Janku said.
Dr. Janku said that systemic reactions to intratumoral C. novyi-NT treatment need to be better controlled before its use becomes widespread. “The treatment can be done safely at MD Anderson and other major cancer centers, but I don’t think it could be easily done at a local hospital,” he said. “On the other hand, I feel far more confident giving this treatment than I did when the trial started; we’ve learned a lot. If applied correctly, C. novyi-NT can be a very powerful tool.”
As with bacterial infections, for many years there were case reports and other anecdotal evidence of patients whose cancer went into spontaneous remission after exposure to or vaccination against a virus. But studies done in the 1960s using adenoviruses to treat cancer showed limited efficacy and virus-related complications.
Little further research was done in this area until the 1990s, when it became possible to genetically manipulate a virus in the laboratory. It was then that Juan Fueyo, M.D., a professor in the Department of Neuro-Oncology, and Candelaria Gomez-Manzano, M.D., an associate professor in the Department of Neuro-Oncology, began working to develop an adenovirus that acts against cells that lack retinoblastoma tumor suppressor protein (Rb).
“We made a virus that cannot inactivate Rb, which prevents the virus from replicating in a normal cell,” Dr. Fueyo said. “But when it infects a cancer cell, Rb is already gone, and the virus can replicate.” In vitro studies revealed that some cells were resistant to the virus, so Drs. Fueyo and Gomez-Manzano modified the virus with the Arg-Gly-Asp tripeptide (RGD) motif because RGD integrins are expressed at high levels on cancer cells but low levels on normal cells.
The resulting virus, Delta-24-RGD (also called DNX-2401), was recently investigated in a phase I dose escalation trial in which it was injected intratumorally in patients with recurrent glioblastoma. The trial’s principal investigator, Frederick F. Lang, M.D., a professor in the Department of Neurosurgery, said, “We escalated all the way up to the top dose, which is 5×1010 viral particles, with no dose-limiting toxic effects.” Among the 25 patients treated, the tumors disappeared completely in three and partially regressed in one.
“When a patient’s glioblastoma comes back, survival often is measured in weeks,” Dr. Lang said. “But these four patients lived more than 3 years after their treatment—and one patient lived 4 years—before their cancer recurred. Three of the patients are still alive.”
Drs. Fueyo and Gomez-Manzano theorize that Delta-24-RGD triggered an immune response. “When you inject the virus, it replicates in the tumor for a short period, but after a while the immune system will identify the virus and destroy it,” Dr. Fueyo said. “But in a few patients—these three whose tumors regressed—we believe that the immune system shifted from targeting the virus to targeting the tumor cells.” Upcoming trials will test whether combining Delta-24-RGD with other immunotherapy drugs will improve patient outcomes.
A phase IB study (2014-0488) of intratumoral Delta-24-RGD with or without subsequent interferon-gamma is now enrolling patients with recurrent glioblastoma or gliosarcoma at MD Anderson and other centers. Another study is being planned that will combine Delta-24-RGD with a PD-1 inhibitor.
Another oncolytic virus, talimogene laherparepvec (T-VEC), which was recently approved for the treatment of advanced melanoma, also will be combined with an immune checkpoint inhibitor in an upcoming clinical trial. The trial is scheduled to begin enrolling patients with melanoma at MD Anderson and other centers later this year.
Procedures for intratumoral injection
In addition to establishing the safety and feasibility of Delta-24-RGD treatment, the phase IB trial of the oncolytic virus helped perfect the procedure for injecting the virus into brain tumors. The resulting procedure uses magnetic resonance imaging (MRI)–guided stereotactic implantation of a new microinfusion cannula developed specifically for injecting therapeutics into the brain.
“We put a frame around the patient’s head, which gives us an XYZ coordinate system, and then we do an MRI with that frame in place to provide points of reference,” Dr. Lang explained. With the patient under local anesthesia with conscious sedation, a hole is drilled in the skull through which, with MRI guidance, Dr. Lang passes the cannula (which is less than 2 mm in diameter) into the brain using an arc-shaped stereotactic device that sits on the frame. “We use this stereotactic system to precisely position our cannula within the tumor,” he said.
Once the cannula is in place, Dr. Lang infuses gadolinium contrast medium to allow him to monitor the virus’s distribution as it pushes the medium along. Then the 1-mL dose of Delta-24-RGD virus is infused at a slow rate (0.9 mL/h) using a microinfusion pump. “The idea of this controlled slow infusion, called convection-enhanced delivery, is that it opens up the interstitial spaces within the tumor by creating a continuous, very slow pressure gradient, and the fluid and virus can move through that interstitial space,” Dr. Lang said.
Compared with the infusion of the oncolytic virus into brain tumors, the injection of C. novyi-NT or dendritic cell vaccine into tumors in other parts of the body is a straightforward procedure.
“The technique is a natural extension of the technology we use every day for image-guided biopsies, but instead of just taking tissue, we’re injecting therapy,” Dr. Murthy said.
The location, size, and characteristics of the tumor determine whether ultrasonography, computed tomography, or (rarely) MRI is used to guide the injection. The type and size of the needle are determined by the tumor’s size and location and the type of therapy to be injected. For example, Dr. Murthy said, “For C. novyi-NT we use a multipronged array needle to ensure even distribution. Because tumors are heterogeneous, we don’t know the best area to target. So we spread the bacteria evenly within the tumor because we think this will give the bacteria the best chance of germinating. The procedure is simple—it just needs to be done right.”
Because dendritic cell vaccine, C. novyi-NT, and Delta-24-RGD are either known or theorized to stimulate the immune system, combining them with immunotherapy drugs was the next logical step in their clinical development.
Dr. Subbiah hopes that combining an intratumoral vaccine and an immune checkpoint inhibitor will enhance the efficacy of both drugs. “Checkpoint inhibitors have changed the landscape of immunotherapy for many cancers,” he said. “But some patients do not respond to these drugs, and others develop resistance. The lack of response may be due to a lack of preexisting immune response in these patients. We all know that dendritic cells are needed for induction of an adaptive tumor response, which is one reason I think the intratumoral dendritic cell vaccine has the potential to augment checkpoint inhibitors.”
In addition to combining Delta-24-RGD with an immune checkpoint inhibitor, Drs. Gomez-Manzano and Fueyo, along with Carlo Toniatti, M.D., Ph.D., executive director of the Oncology Research for Biologics and Immunotherapy Translation platform in MD Anderson’s Institute for Applied Cancer Science, and Laura Bover, Ph.D., an associate professor in the Department of Genomic Medicine, hope to go one step further. “We are trying to make a virus that itself will express a positive regulator of the checkpoints,” Dr. Fueyo said. The new virus, Delta-24-RGDOX, expresses the ligand for OX40 (also called tumor necrosis factor receptor superfamily member 4), which is expressed on tumor-specific T cells. Delta-24-RGDOX has shown promise in preclinical studies. “We are at the interface between virotherapy and immunotherapy,” Dr. Fueyo said.
In addition to its use for delivering vaccines and oncolytic viruses and bacteria, intratumoral injection may provide the answer in areas of immunotherapy research that seem to have stalled. “Many of the treatments we give to stimulate the immune system, such as toll-like receptor antagonists, never actually make it into the tumor and are too toxic when given systemically,” Dr. Murthy said. Indeed, a trial of an intratumorally injected toll-like receptor antagonist is in the planning stage.
Another upcoming trial will test intratumoral injection of avidin followed by an intravenous injection of radioactive biotin, which binds irreversibly and with an extreme affinity to avidin. This trial is being planned by Drs. Murthy and Subbiah along with Gregory Ravizzini, M.D., an assistant professor in the Department of Nuclear Medicine. Still more trials of intratumoral oncolytic treatments are expected to follow.
“These intratumoral treatments are blossoming,” Dr. Murthy said. “It’s an exciting time to be involved in this area of clinical research.”
Dr. Janku agreed. He said, “These treatments are making their way into the standard of care.”
For more information, contact Dr. Juan Fueyo at 713-834-6221, Dr. Filip Janku at 713-563-0803, Dr. Frederick F. Lang at 713-792-2400, Dr. Candelaria Gomez-Manzano at 713-834-6260, Dr. Ravi Murthy at 713-745-0856, or Dr. Vivek Subbiah at 713-563-0393. To learn more about ongoing clinical trials at MD Anderson, visit www.clinicaltrials.org.
OncoLog, May 2016, Volume 61, Issue 5