Immune cells programmed to attack tumours in a smarter way have shrunk brain and ovarian tumours in mice studies where unaltered immune cells failed. The technology could be used to treat cancers as well as degenerative brain disorders.
“We have more control over what the cell does when it reaches the disease site,” says Kole Roybal at the University of California, San Francisco. “We can really program in very specific functions.”
Our bodies naturally kill off many nascent cancers, but sometimes immune cells called T-cells don’t recognise cancerous cells. One way to treat cancers that manage to dodge the immune system is to genetically engineer T-cells to produce a receptor that helps them target a specific protein on the cancer cell’s surface. These are called CAR T-cells, where CAR stands for chimeric antigen receptor.
CAR-T therapies have cured a few people, leading the US to approve two forms in 2017. But there are major limitations. The approach has only been effective against blood cancers such as leukaemia, not against solid tumours. And it can have very serious – even fatal – side effects if the T-cells kill off non-cancerous cells that also have the target protein on their surface.
These problems are related. One of the reasons why CAR-T therapies don’t work for solid tumours is that not all cells in such tumours express a single, unique protein, says Roybal. So his team has developed a new type of receptor protein that works in a different way. Instead of triggering an instant attack, these T-cell receptors switch on any desired gene or genes when they recognise a target protein. This can be any protein the researchers choose, which is why the technique can be used for brain disorders in addition to cancers.
Read more: Gene editing beat a baby’s leukaemia. Are other cancers next?
Roybal’s team engineered this receptor to recognise a protein specific to some cells in brain tumours called glioblastomas. The receptor then activated a gene for a standard CAR-T receptor that targets a protein found on a wider range of tumour cells and on healthy cells. Crucially, though, the killing effect was limited to tumour environments where both proteins are present: if the engineered cells leave the tumour, the CAR-T gene gets switched off again.
In tests in mice, this approach shrunk glioblastomas and prevented recurrence where conventional CAR-T therapies either didn’t work or didn’t prevent regrowth. In a separate animal study, similar results were found for ovarian cancers and mesotheliomas, which are mostly caused by asbestos.
Standard CAR T-cells seem to become exhausted relatively quickly and die off, says Roybal. The smart CAR T-cells persisted for longer in the body, which is important for preventing recurrence, he says.
“We are solving a load of the roadblocks in solid tumours,” says Roybal. “We are not all the way there, there’s a lot of work to do, but we have taken major steps.”
For instance, tumours often release factors that suppress an immune response. His team plans to engineer the smart CAR-T cells to release other factors that counteract this, and stimulate a broader immune response against tumours.
His team is also getting the therapies used in the mice ready for human trials, which could take a couple of years. These trials will involve removing a patient’s immune cells and genetically engineering them before putting them back in the body. In the longer term, it may be possible to treat people using “off-the-shelf” cells, which would greatly reduce costs.
Journal References: Science Translational Medicine, DOI: 10.1126/scitranslmed.abe7378 & 10.1126/scitranslmed.abd8836