Washington - People facing one of medicine's most dreaded diagnoses may someday have an unlikely source of hope: Injections that treat some types of brain cancer with part of a deadly tropical virus.
The virus, called Lassa, is the cause of a particularly nasty hemorrhagic fever endemic to West Africa. The research sounds terrifying, but a team from Yale and Harvard universities is using part of Lassa's genetic code to make another virus safe enough to inject into the human brain, where it will hunt down and kill cancer cells.
Lassa's partner in this search-and-destroy mission is vesicular stomatitis virus, or VSV. Most people who contract it never get sick, and those who do usually have only flulike symptoms. But if it infects the brain, VSV becomes deadly. Ironically, it's the more fearsome Lassa that makes the usually innocuous VSV safe for use in that delicate organ.
Some viruses, including VSV, are good at attacking cancer cells without infecting healthy cells nearby. Many types of cancer cells are vulnerable to certain viruses, including VSV, because they can't produce interferon, a protein that normal cells use to fight off infection. Some researchers see these viruses as the solution to one of their biggest challenges: killing the cancer without harming the patient.
Researchers have tested VSV against several kinds of cancer in lab experiments, and it's usually effective without damaging healthy cells. A Mayo Clinic team started a clinical trial in 2012 to test VSV against liver cancer; it's still underway. “We've been working with VSV for a number of years, and VSV has always been a really good virus for infecting cancer cells,” said Anthony Van den Pol, a Yale neurosurgery professor whose team has tested VSV against cancer in lab animals, as well as in cultures of human cells. “But the problem has always been that if VSV gets into the brain, it can cause neurological problems or even death.”
In the brains of lab mice and in cultures of human brain cells, VSV attacks cancer, but it also kills healthy brain cells. It seems that VSV's glycoprotein - a protein on the outer surface of the virus that latches onto host cells - binds with neurons too easily. Once that happens, the virus takes over the cell and uses the cell's internal structures to produce new copies of the virus. Eventually, the infected cell will burst, scattering virus particles like shrapnel to attack neighbouring cells.
Although neurons' long tails, called axons, carry nerve impulses all around the body, their bodies, or soma, are primarily found in the brain and spinal cord. It's generally safe to use VSV against cancer in most parts of the body, because a barrier usually keeps viruses in the bloodstream from reaching the brain and spinal cord, but it's too dangerous to use the virus against brain cancer.
Each virus' glycoprotein is shaped differently, so Van den Pol and his team thought that if they could replace VSV's glycoprotein with one from another virus, it might be safe enough to use in the brain. The list of candidates was a roll call of frightening viral diseases: Ebola, Marburg, lymphocytic choriomeningitis virus (LCMV), rabies and Lassa. These five viruses had glycoproteins that, like VSV'S glycoprotein, would bind to a broad range of cell types, so the researchers thought there was a good chance that they would also bind to several types of cancer.
Van den Pol and his colleagues did not work with live samples; rather, they used sections of each virus' genetic material. The team removed the gene that coded for VSV's glycoprotein and replaced it with a gene coding for another virus' glycoprotein. When the team was done, it had five new, man-made viruses. Of the five, the most effective was Lassa-VSV.
The Lassa-VSV virus is a chimera, an organism that's created by combining the genomes of two different organisms to make something new. It kills brain tumours without damaging healthy neurons, apparently because it can bind with normal cells but doesn't actually infect them.
“We think that normal cells are able to block the replication of the virus, and the virus doesn't get a foothold,” Van den Pol said. The team published its results in the Journal of Virology.
The researchers tested their chimeras against some of the deadliest forms of brain cancer. Glioma can kill people within a year. The prognosis can be even worse for patients with the skin cancer melanoma if it spreads into the brain.
In lab cultures of glioma and melanoma, Lassa-VSV and Ebola-VSV worked so well that the researchers decided to test them on mice. They implanted human glioma cells in the mice's forebrains. Animals that didn't receive treatment died within about 29 days. The mice that got Lassa-VSV injections were doing fine 80 days later, when researchers declared them survivors. The virus had destroyed the tumours without damaging the brain cells. Those that received Ebola-VSV fared slightly less well but still better than those that got no treatment.
In high-grade gliomas, cancer cells often migrate within the brain. When that happens, surgery or radiation might miss a rogue group of cells, leaving the cancer an additional foothold. When researchers implanted two tumours into mouse brains, one in each hemisphere, and then injected the mice with Lassa-VSV, the virus infected and killed one tumour, then diffused through the brain to attack the second one without infecting brain cells along the way.
However, Lassa-VSV won't work every time. Every cancer is caused by a unique set of mutations. Some of those mutations make tumours more susceptible to Lassa-VSV, but others make the cancer a harder target for the virus.
For example, Lassa-VSV was less effective at killing lab cultures of one type of glioma, U118, than another type, U87. Other chimeras might be better able to target those viruses, however. The Ebola-VSV chimera used in Van den Pol's study was effective against U118, albeit less so than Lassa-VSV, and a 2008 study found that the LCMV-VSV chimera also had the potential to safely hunt down brain tumours.
Successful treatment in mice doesn't guarantee that it will work in humans. “We hope that what we learn from rodent studies can generalise to humans, but of course there is the possibility that there is some difference in humans that might result in less or no efficacy of Lassa-VSV,” Van den Pol said in an email.
Primate research would usually be the next step. “The first thing we would do is inject the virus into a primate brain with the primary question being 'Is this safe or not?' because if it's not safe, it's the end of the road,” Van den Pol said. Such testing requires a clinical-grade, uncontaminated version of the virus, produced in a special laboratory.
It's an expensive process. “In my lab, I can make a liter of clean virus for about $100 (about R1 000) or $200, but to get a clinical-grade virus would cost about a million dollars, and I just can't find the money,” Van den Pol said. It can be difficult to find anyone willing to fund potential treatments at this stage, he said, because there is still a chance that Lassa-VSV won't work in humans. “You know, if it works, it's almost nothing, and if it doesn't work, of course, it's a large investment in nothing.”
Meanwhile, researchers are testing Lassa-VSV against melanoma in mice. The team is especially interested in whether the virus can not only kill cancer cells but also help the patient's immune system target and destroy cancer cells.
As part of the body's immune response, T-cells recognise and destroy infected cells before they can release more virus particles and spread the infection. Some cancer cells produce molecules that trick the immune system into ignoring them, but when Lassa-VSV infects a cancer cell, it may signal T-cells and other immune system cells to attack the infected cell.
By targeting cancer cells infected with Lassa-VSV, the body's immune response may also learn to recognise and target cancer cells long after the virus is gone. The Lassa-VSV virus carries a gene that causes it to produce a fluorescent protein that helps researchers track which cells have been infected by the virus. In mice, the immune system produces antibodies to target not only the virus but also the protein. Researchers hope that if the immune system can learn to target a protein associated with the virus, it might also generalise from infected cancer cells to all cancer cells.
If it works, that could help protect patients against a relapse. “If the tumour cells were to return X number of months later, then the immune system would still recognise the tumour as something to target,” Van den Pol said. Researchers still have some work to do to understand how immune cells inside a tumour differ from normal cells.
Currently, the team is studying how Lassa-VSV works against melanoma tumours in mice, and Van den Pol said the results are promising. Mice that had Lassa-VSV injected into their tumours “have substantially smaller tumours and are still doing fairly well at this point,” said Van den Pol, although he added that the team isn't sure yet whether the body's immune system will take over and eliminate the melanoma completely or whether it will just keep the tumour in check but not destroy it.
Van den Pol is optimistic about Lassa-VSV'S potential. “The fact that Lassa-VSV targets both glioma and melanoma suggests that the virus could be good at targeting many other types of cancers,” he said.
* Smith-Strickland is a freelance science journalist based in Kansas. Follow her on Twitter: @RescueFins.
The Washington Post