Advanced – But Not Advanced Enough

The gene therapy industry has been building up steam for many years – and it’s hard to say if there was a specific turning point that signified the start of the current era of success. Ultimately, I think much of what we’ve achieved in this field so far comes down to perseverance. When I was working in the field early on, the main discussion points were retroviral vectors and severe combined immunodeficiency (SCID) – sometimes known as “boy in a bubble” syndrome because the normal environment can often be fatal. SCID can be treated with a bone marrow transplant (usually hematopoietic stem cell transplant), but not all patients can be matched with a donor. One famous SCID case is that of David Phillip Vetter, born on September 21, 1971, who spent most of his life in a sterile, protective bubble. No donor match could be found, but as bone marrow transplant science improvedd, it became possible to use donors who weren’t an exact match. David received a transplant from his sister in 1983, but the bone marrow contained trace amounts of Epstein-Barr. The virus evaded pre-transplant screening and David died from a form of lymph cancer on February 22, 1984.

SCID was an early research avenue for gene therapy; however, early gene therapy attempts were somewhat crude and made use of retroviruses – and there were concerns that using a retrovirus to drop in a gene could trigger other diseases, such as cancer. The problems prompted the scientific community to look at the biology of retroviruses and find ways of ensuring that a given gene could be inserted without causing unwanted effects.

For a while, the industry was very interested in adenoviruses as a gene therapy vector, but these were not very successful and caused some safety concerns – most famously with the Jessie Gelsinger case. That said, adenovirus vectors have been very good for vaccines – as we’ve seen with COVID-19. Adenovirus-based gene therapies have also been approved in China, but other countries have not followed suit.

Eventually, there was a jump to the adeno-associated virus (AAV) as a vector. It was considered a niche area at first but people kept digging away at the science – and eventually it led to approvals for Luxturna and Zolgensma. The approvals of these therapies were significant milestones for the gene therapy industry and have catalysed growth in the field. Yes, there are still great challenges that the gene therapy field faces – costs are high and these therapies can come with significant risks to patients. However, I can’t see us going backwards!

Empty or full?
 

There are several areas where gene therapies can be improved. AAV is the most common vector used today for in vivo gene therapy, but around 50 percent of the population have pre-existing immunity to AAV, which means they are unlikely to benefit from the therapy. If we can screen these people, they can be excluded from trials or alternatively we could look at ways of increasing the dosing.

Another key area of focus for AAV vectors is the ratio of empty capsids to full, and whether this affects toxicity and immunogenicity. In September 2021, the FDA held a meeting about dose and toxicities – with capsids being a key discussion point. Full capsids are required for therapeutic efficacy, but it’s not known if empty capsids are a problem. However, because they do not contribute to efficacy (although some researchers believe there could be some benefits to empty capsids) there is a logical argument that we should reduce them.

A virus’ natural replication system is very efficient, so using a wild type adenovirus to help produce an AAV would result in lots of full AAV capsids. However, once scientists begin to engineer the virus – which is often required for drug development – the replication efficiency reduces and the yield of full capsids decreases. Science and understanding in this area will continue to evolve, but we do know that mixing elements from different serotypes or tampering with the biology has a negative impact. In short, the more natural the things are, the better the yields appear to be. 

Many scientists – and companies – are working to find solutions to this capsid issue. But the concern here is that whoever finds a good answer is more likely to patent it than share the information – after all, that is the nature of the pharmaceutical industry!

A wish list of new technology
 

There are many other challenges that we must address in gene therapy. In some ways, we can apply lessons learned from mAbs because the process of making mAbs and viral vectors is quite similar. A key technology for viral vectors is likely to be single use equipment – and the mAb area can teach us a great deal about these systems. Fixed stainless steel for vectors may come, but there are too many variables in the gene market right now for companies to consider such a rigid manufacturing infrastructure, so single use will likely be the way forward for the immediate future.

However, we can’t take all of our knowledge from mAbs. Viral vectors are made using plasmid transfection, which can have scale limitations. Using a 1000 L or 2000 L bioreactor for viral vectors will improve throughput and costs, but the volumes of plasmid and transfection reagents, as well as mixing at large scale, are significant challenges (e.g., cost, process reproducibility). New technology in this area will definitely be helpful. In addition, we could also benefit from improvements in the price of plasmids – they are very expensive for what they are, especially for GMP grades!

Cell lines for gene therapy also face issues. With mAbs, you can generate a cell line within a couple of months; but this isn’t possible when it comes to cell lines for a gene therapy vector. Making cell lines for gene therapy vectors is a huge effort because we’re using 30-year old technology that controls gene expression – the cells do not like it when you constitutively express certain viral genes, even at a low level. Right now, most companies are using the transient transfection route because it’s flexible and you don’t need to invest in a cell line for a product that may not eventually proceed to market. However, transient transfection can be limited in terms of cost, yields, and scalability. We need new approaches and technologies that can more quickly generate cell lines for viral vectors.

Purification represents another opportunity for progress. With antibodies, you are somewhat spoilt for choice when it comes to purification, with a large range of chromatography resins and filters available. And there is a lot of expertise to go around. But for the purification of viral vectors, the technology offering right now is limited. Fortunately, progress is being made; for example, Thermo Fisher Scientific, Cytiva, and Repligen are all active in this space, though there is often a reliance on users to help bring new technologies to market, which can be frustrating for users. I’d like to see more companies putting out solutions that are truly ready for use.

Finally, I would like to see more work in the area of analytics. Right now, we don’t understand the products very well or what methods we need. I like to use the analogy of nailing jelly to the wall! It seems as soon as you have qualified and verified the right analytical method, people start questioning whether it is actually the right method, or a “better” method comes along. Each method also seems to have its pros and cons; for example, analytical band centrifugation does a good job of determining your empty:full capsid ratio, but you need a lot of material and you can only perform the analysis at the end of the process.

If you want to follow your empty:full ratio throughout the process then there are very few methods available. You can use cryo-electron microscopy, but this is expensive, very specialized, and the results take weeks. There is also capillary ion exchange chromatography, which can potentially separate empty capsids from full capsids, but it can't do partials. The equipment is also expensive and if the assay gets registered in a filing – and then the technology moves on – you are stuck with the equipment.

Are empty:full capsids going to be important a few years from now? We don’t know – so investment decisions are being made based on incomplete information, with the industry chasing its tail to catch up to the next thing that comes along in analytics. I’m pleased to say there is at least a lot of transparency in this area; many different companies are pushing different technologies, which is leading to a lot of discussion. Ultimately, this activity should help decrease costs and lead to more options coming onto the market. 

Right now, the testing landscape can be confusing, with companies left with questions about what technology is right for them and whether they should wait another year to see if anything changes.

In it together
 

How do we move forward? I would like to see the industry working more collectively to advance the gene therapy field – product developers, CDMOs, equipment companies, and regulators. The mAbs field has been growing since the 1990s – with huge strides made in the 2000s. Gene therapy has a way to go, but we can get there. One significant call to action is knowledge sharing. Right now, this is not where it should be. The industry goes into research holes to solve issues and then somebody pops up with a solution, but the discussion on what has been learned and what it means is not as transparent as it should be. It can be hard to know what’s going on and who’s working on what, which leads to many people working in parallel on the same problems and making the same mistakes – a real waste of resources. 

It’s understandable that companies want to maintain a competitive edge, but the lack of sharing and transparency is not helping the industry as a whole. Slowly, we will see change – especially as more partnerships are formed between developers and their CDMOs and equipment companies, and also via industry forums such as BioPhorum and ARM. It would be naive to think we won’t hit more bumps along the way, but we’ll be able to get over them faster and easier as a collective.