Developing new treatments for intractable cancers requires outstanding minds working together toward a common goal
By Darrell E. Ward
Photo illustration by Roman Sapecki
Oncolytic viruses—viruses that are engineered to kill cancer cells—offer a promising strategy for treating cancers that remain steadfastly incurable. That is, if the virus can get a toehold in enough tumor cells to work.
When oncolytic viruses entered clinical testing in the late 1990s, they showed no evidence of harm but little evidence of good. “The viruses were safe but ineffective. They were too attenuated,” says E. Antonio Chiocca, MD, PhD, professor and chair of Neurological Surgery at Ohio State, Dardinger Family Endowed Chair in Oncological Neurosurgery and a leader in the field of oncolytic virus therapy for brain tumors.
Subsequent research showed that host defenses and changes in the tumor microenvironment quickly eliminated the emasculated viruses before they had a chance to kill tumor cells.
Chiocca’s interest in oncolytic virus therapy began during his residency at Massachusetts General Hospital, where he spent two and a half years working in the laboratory of his mentor, Harvard’s Robert L. Martuza, MD, FACS.
“Bob was trying to learn if viruses could be used to kill cancer cells,” Chiocca says. “We began collaborating with the herpes virus group at Harvard, and then I was hooked.”
Chiocca, who is co-leader of the OSUCCC-James Viral Oncology Program, specializes in malignant gliomas, particularly glioblastoma multiforme, which has a median survival of about 15 months. These tumors, with their multiple extensions into the surrounding gray matter, are difficult to cure even with surgery, chemotherapy and radiation, Chiocca says. But they are excellent candidates for oncolytic virus therapy.
Since arriving at Ohio State in 2004, Chiocca’s research has focused on developing oncolytic viruses based on herpes simplex virus type 1 (HSV-1) that are more potent and more selective for tumors of the brain, and on understanding the host responses that limit or impede the virus’s ability to destroy the tumor.
To ask the right questions and generate imaginative hypotheses related to tumor targeting, intracellular antiviral defenses, inflammatory and innate immune responses, angiogenic changes and extracellular matrix alterations, requires talented collaborators.
“Science has become more complex,” Chiocca says. “Work such as ours requires people with different insights, skills, ideas, points of view and expertise…collaboration can make research more relevant.”
Chiocca works with a number of accomplished collaborators, and together they form an oncolytic virus group.
Working with these colleagues, Chiocca has developed several oncolytic herpes strains, two of which—MGH2 and rQNestin34.5—he hopes to test soon in phase I trials. “Both are more potent and more selective for tumor cells, as well as being safe,” Chiocca says.
The rQNestin34.5 oncolytic virus exemplifies Chiocca’s progress. This oncolytic virus is attenuated in two ways. First, the gene encoding viral ribonucleotide reductase is deleted from the viral genome, leaving the virus unable to replicate without an outside source of the enzyme. That outside source is provided by tumor cells with defects in the p16 tumor-suppressor pathway. These cells, even when quiescent, overexpress the enzyme and so enable viral replication.
The second novel attenuation involves the viral ICP34.5 neuro-virulence protein, which normally blocks a defense mechanism that pushes infected cells into apoptosis. Wild type HSV-1 has two copies of the gene, both of which were deleted from early HSV-1 oncolytic virus mutants to ensure safety.
To improve potency in the rQNestin34.1, Chiocca and his colleagues restored one of the two ICP34.5 genes. To maintain safety, the researchers coupled the viral gene to a glioma-specific promoter. The promoter comes from the gene for the protein nestin, an intermediate filament in embryonal brain cells.
Chiocca and colleagues showed in a 2005 study published in the journal Cancer Research that nestin is expressed in malignant adult glioma cells but not in healthy adult astrocytes.
Furthermore, their in vitro studies suggested that combining the nestin promoter with ICP34.5 improved tumor specificity by enabling viral replication in glioma cells but not in healthy astrocytes, and it made the virus more cytotoxic to glioma cells. In mice with established gliomas, the nestin virus increased long-term survival by 50 percent when administered early as therapy, and it significantly increased survival even in symptomatic animals.
More recently, Chiocca has identified a population of tumor- initiating cells that is resistant to his HSV vectors. Working with Chiocca, herpes virus specialist Joseph C. Glorioso III, PhD, at the University of Pittsburgh School of Medicine, is developing vectors with tumor-specific receptors that target those resistant cells. Glorioso and his laboratory also grow and purify the vectors needed by Chiocca for preclinical studies that use human brain tumors in animal models. Of major interest are the creation of vectors that target epidermal growth factor receptor found on glioblastoma and tumor stem cells, and the engineering of oncolytic HSV that depends on the differential expression of microRNAs for selective replication in brain tumors.
While Chiocca prepares the rQNestin34.5 virus as therapy for adults with brain tumors, one of his collaborators, pediatric oncologist Timothy P. Cripe, MD, PhD, at Cincinnati Children’s Hospital Medical Center, is studying its potential for treating neuroblastoma in children.
“We have developed some very good therapies for these tumors over the years, but we are unable to cure the majority of patients with metastatic disease, so there is still a large unmet medical need,” Cripe says.
“Our collaboration with Dr. Chiocca has been very fruitful. We have found that the nestin protein is upregulated in neuroblastoma, suggesting that this tumor may be another target for the rQNestin virus,” he says.
“We’re also exploring the addition of cyclophosphamide to suppress the immune response to oncolytic virus infection, something that Dr. Chiocca pioneered,” Cripe says.
That finding came from a 2006 study led by Chiocca and published in the journal Proceedings of the National Academy of Sciences. The researchers showed that high numbers of innate immune cells—natural killer (NK) cells, macrophages and microglia—are drawn to the tumor within six hours of injecting an oncolytic virus.
They found that the drug cyclophosphamide briefly dampens this immune cell activity, giving the oncolytic virus more opportunity to disperse through the tumor and kill cancer cells. Animals given the drug and an oncolytic virus showed a 50- percent increase in the number of tumor macrophages, for example, compared with a three-fold increase in animals given the virus and no drug.
“Cyclophosphamide seems to temporarily inhibit just this early immune response, making it unnecessary to totally suppress the immune system during treatment,” says Chiocca.
The agent inhibits the innate immune response at least in part by inhibiting production of interferon gamma (IFN-g) by NK cells, according to the findings of work reported in the same paper and led by Michael A. Caligiuri, MD, an authority in NK cell biology, director of Ohio State’s Comprehensive Cancer Center and CEO of the James Cancer Hospital and Solove Research Institute.
IFN-g attracts immune cells to an infection site, which could intensify the immune response against the anticancer virus. In rats with brain tumors that were treated with the virus alone, IFN-g levels rose by a factor of 10 after six hours, and by more than 120 times after 72 hours. However, IFN-g levels rose only slightly in animals treated with the virus plus cyclophosphamide.
Overall, the study suggests that cyclophosphamide can improve oncolytic virus therapy by delaying the activity of NK cells and other innate immune cells.
Seeking NK details
NK cells have both antiviral and antitumor properties, so it is crucial to understand how they interact with tumors infected with an oncolytic virus. Christopher A. Alvarez-Breckenridge, a student in Ohio State’s combined MD/PhD degree program and a member of Chiocca’s lab, is identifying the signals exchanged between NK cells and infected versus uninfected tumors.
“Our preliminary findings indicate that NK cells are moderately activated by uninfected brain tumors in an animal model but highly activated by virus-infected tumors,” says Alvarez-Breckenridge, who works closely with the Caligiuri laboratory. “And if we culture NK cells together with uninfected and infected tumors, they preferentially clear viral infected tumors.”
Overall, Alvarez-Breckenridge says, his findings suggest that suppressing the immune system for 72 hours, perhaps less, might give the virus sufficient time to replicate and spread. “At that point, the immune system might actually aid the virus and help clear the tumor,” he says.
Chiocca collaborator and cancer center member Balveen Kaur, PhD, associate professor of Neurological Surgery, studies how the tumor microenvironment impedes oncolytic virus therapy for gliomas.
An animal study led by Kaur found that, three days after an oncolytic virus is injected into a tumor, tumor blood vessels become leaky, immune cells infiltrate the site, and IFN-g expression rises. Administering one dose of the angiogenesis inhibitor cRGD (cyclic peptide of arginine-glycine-aspartic) prior to viral treatment, however, reduced vessel permeability, immune cell infiltration and IFN-g expression.
In addition, rats treated with the agent had higher viral titers and significantly lower tumor vasculature (28 vessels versus 62 in control animals per area of viewing field), and the treated animals showed a 23- percent increase in survival.
“The survival increase was significant because these are very aggressive tumors,” Kaur says. “This work suggests that antiangiogenic agents can reduce virus-induced inflammation in brain-tumor tissue and improve the efficacy of oncolytic virus therapy by slowing the immune system’s ability to clear the virus.” The 2007 study was published in the Journal of the National Cancer Institute.
The following year, Kaur and her colleagues found that, as oncolytic viruses destroy glioma cells, the cells release proteins that stimulate tumor angiogenesis. The study, published in the journal Molecular Therapy, found that virus-treated tumors had roughly five times more blood vessels than untreated tumors. These vessels facilitate the innate immune response that eliminates the virus, and they support re-growth of residual tumor cells.
Probing further, Kaur and her colleagues identified three genes linked to blood-vessel growth that were overexpressed, in particular. Of those, the gene called CYR61 was nine times more active in virus-treated tumor cells than in uninfected tumor cells, and the activity increased in a dose-dependent manner.
“This change in gene activity may represent a general host response to the viral infection,” says Kaur.
The investigators are now studying whether expression of this gene might serve as a biomarker reflecting patients’ response to oncolytic virus therapy. “Measuring a patient’s res-ponse to viral infection is currently not feasible,” Kaur says, “so if this were to work, it would be a signi-ficant advance.”
Kaur’s findings have led her and her colleagues to modify the oncolytic virus by adding a gene for a natural, brain-specific angiogenesis inhibitor called vasculostatin. They called the new virus RAMBO, for Rapid Antiangiogenesis Mediated By Oncolytic virus.
The researchers tested the new virus in six animals with dermal xenografts of human glioblastomas. RAMBO-treated animals survived an average of 54 days, and three mice were tumor-free at the end of the experiment. Control animals, treated with a virus that lacked vasculostatin, survived an average of 26 days, and none were tumor-free.
When they treated five animals with a human glioblastoma in the brain, the animals survived 54 days, with one remaining tumor-free for more than 120 days.
A time-course study showed that, after an initial period of shrinkage, implanted brain tumors began regrowing about day 13 in control animals, but not until about day 39 in RAMBO-treated animals.
The study, published in Molecular Therapy in 2010, “shows a significant antitumor effect and supports further development of this novel virus as a possible cancer treatment,” Kaur says.
Chiocca’s rQNestin34.5 and MGH2.1 oncolytic viruses, both second-generation agents, are undergoing preclinical testing in preparation for possible phase I trials in people with malignant brain tumors. Meanwhile, he and his colleagues are moving forward with third-generation viruses such as RAMBO and the development of oncolytic virus therapy for children.
“The atmosphere here at Ohio State encourages cooperative research,” says Chiocca. “People check their egos at the door for the common good. Having a first-rate team of investigators allows us to do stronger work, write better papers and higher quality proposals, compete for higher quality journals and obtain better funding.”
Outstanding minds coming together toward a common goal can offer real hope to people with an intractable cancer. It is hope born of collaboration.
AN ONCOLYTIC MEASLES VIRUS
Corey Raffel, MD, PhD, professor of Neurosurgery at Nationwide Children’s Hospital in Columbus, a close collaborator of E. Antonio Chiocca, is developing oncolytic virus therapy for children with meduloblastoma using a modified measles virus. About 350 meduloblastoma cases occur annually in the United States. The malignancy has about a 70-percent survival rate. “The problem is that curing the tumor requires radiation therapy, which is not good for a child’s brain, so the kids who survive often do poorly,” Raffel says. “We are trying to find ways of curing them without the devastating effects related to radiation therapy.” He is studying the use of an oncolytic measles virus with a receptor that is aberrantly upregulated on meduloblastoma cells, enabling the virus to target and enter the tumor cells. The altered measles virus has been tested in adults but not in pediatric tumors. Raffel began his oncolytic virus work about five years ago while still at the Mayo Clinic. One of the two animal models he has developed enables him to study disseminated meduloblastoma, which occurs when the tumor spreads through the cerebrospinal fluid to another location in the central nervous system. Raffel attends many of Chiocca’s weekly lab meetings to exchange information, discuss data and suggest ways to improve experiments and overcome problems. “We work with different viruses, but these meetings offer valuable opportunities to show your work to others and receive comments and suggestions,” Raffel says.