Daniel Sze (Stanford University, Stanford, USA), presenting at the European Conference on Interventional Oncology Annual Meeting (ECIO; 22–25 April, Vienna, Austria), calls for a synergy between gene therapy and interventional oncology, saying that delivery efficiency can be improved by interventional oncologic procedures, cutting costs on treatments which currently cost up to US$1,200,000 per patient.
Speaking in the session entitled ‘Immunotherapy in 2018’, Sze tells the ECIO audience: “Even after we have talked about this for 45 years, there are still few approved gene therapies for human disease.” The first gene therapy to be approved was in 2003, when the drug Gendicine was granted clearance in 2003 to treat head and neck squamous cell carcinoma (HNSCC). Since then, only a handful of gene therapies have been approved; most recently, in 2017, the US Food and Drug Administration (FDA) granted marketing authorisation to two CAR (chimeric antigen receptor)-T cell therapies: Kymriah (Tisagenlecleucel, Novartis) for B-cell acute lymphoblastic leukaemia and Yescarta (axicabtagene ciloleucel, Kite) for diffuse large cell lymphoma.
“Even though interventional radiology is expensive, gene therapy is even more expensive”, Sze explains. “For instance”, he continues, “those CAR-T cells that were approved last year, Yescarta and Kymriah, cost US$500,000 per patient. Strimvelis [another gene therapy; GlaxoSmithKline] is US$800,000, and Glybera [uniQure] costs US$1.2 million per patient to be treated. In fact, Glybera is going off the market right now, because in the six years since it has been approved, it has only been used to treat one patient, because only one patient could afford the cost.
“So we have to decide, as a field and as a society, where to spend the money. There is the cost of manufacture of these genetic therapies, as well as the cost of administration.”
If, instead of pumping money into the manufacture, Sze argues, the financiers spend slightly more money on the treatment’s administration, then the overall cost of future gene therapy could be significantly reduced. Currently, gene therapy is so expensive in part because scientists need to grow billions or even trillions of vectors to transfect the host cells and express the desired protein product. Transfection can occur either in vivo, where the patient is injected with the viral vector, or ex vivo, where host cells are removed, the viral vector is added, and the transgenic host cells are reintroduced to the patient’s body. Either method requires multitudinous vectors, which all need to maintain the ability to transfect cells, and if transfection occurs ex vivo, sterile, viable human cells need to be grown on a large scale.
“I propose we could spend a little bit more money on the administration. We could perhaps have the manufacturers manufacture a little bit less vector, and we inject it directly into the organ of interest. Likewise, we could have them grow a few fewer cells, and we could inject them into the region of interest,” Sze suggests.
Vector efficiency: “Let the experts have a crack at it”
“Delivery (in vivo); let the experts have a crack at it” reads the slide behind Sze as he advises utilising the skills of interventional oncologists in improving vector efficiency. A literature review in Anticancer Research cites a variety of possible routes for the administration of viral vectors: intravenous, intra-arterial, intra-tumoural, intra-portal, intra-biliary, intra-splenic. Listing these, Sze says, “Of course, this is focusing on the liver, since the liver is the factory of the body. Of all of these, anyone can do intravenous, but these other types of administration are all in the purview of interventional radiology. So the virologists should work on choosing and engineering optimal vectors, the cancer biologists should focus on choosing optimal genes, but the interventional radiologists should work on the optimal delivery.”
Sze and colleagues have looked at some standard interventional radiology practices, and found that many are readily adaptable to gene therapies, and there are even some specificities for tumour cells over normal cells. For instance, Sze used a wedged retrograde hepatic venous delivery in pigs and baboons of a non-viral vector encoding human factor IX, resulting in expression of the human gene, though the expression was not durable.
Writing recently in Cardiovascular Interventional Radiology, Oliver Pellerin and colleagues (Hôpital Européen Georges-Pompidou, Paris, France) provide an example of interventional oncologic methods applied to a gene therapy treatment. They report the successful intra-arterial injection of a rabbit VX2 hepatic tumour with engineered mesenchymal stem cells. These cells are transgenic, possessing a gene that metabolises the prodrug cyclophosphamide 13 times better than the wild type. Following injection with the engineered cells, the investigators administered intravenous cyclophosphamide. While cyclophosphamide alone did suppress tumour growth, the addition of the genetically engineered mesenchymal stem cells helped to metabolise the drug and enhanced efficacy.
Sze concludes: “After nearly 50 years of study, only a few gene therapies are approved for human use, in part because multigenic diseases such as cancer are very challenging. Interventional oncologic methods, however, may improve efficacy and reduce toxicity, as well as reduce costs, due to locoregional delivery of reduced doses.”