The gene therapy field has grown tremendously since 2017, when the first directly administered gene therapy in the U.S., Luxturna, was greenlit by the FDA. In 2018, $9.6 billion was invested globally into the field.

These first-of-their-kind therapies work by delivering modified genes directly into a patient using a vector – in many instances, the naturally occurring adeno-associated virus (AAV) – to replace missing or defective genes or add news ones to treat illness. In the case of Luxturna, the treatment replaces abnormal genes to restore vision loss.

Close to 20 companies are conducting clinical trials for AAV-based gene therapies, with three programs in phase 3 trials. What this means, however, is that facilities are needed to develop treatments at scale.

Here’s what you need to know about gene therapy manufacturing.

The Costs of Manufacturing Gene Therapies

The first considerations are how the virus packages up the DNA of interest, the intended use of the product, and how much virus is needed to get the effective dose.

When thinking about viral production capability, manufacturers should consider the viral particles per liter – basically, the current quoted yield from a bioreactor. When treatments go through purification, there’s approximately a 30% yield, so you can calculate how much you would get per batch. You can then determine how many doses that translates into for a specific application.

Some products have manufacturing costs in the hundreds of thousands of dollars for a large dose indication; others are down to $1,000 for a low dose indication. It depends on the type of treatment being considered: an eye indication would be on the low end, then a brain indication, a major organ would be pricier, and, finally, the whole body. Depending on a patient’s age and weight, that can vary at the disease’s onset.

Then there are facility costs. Effectively, how much time will it take to manufacture that product?

Depending on the scale of manufacturing (whether it’s designed just for clinical production at a small scale) there will be an expenditure per week. Costs could be anywhere from hundreds of thousands to millions of dollars per batch at a decent-size facility.

The Importance of Staffing

Facilities need people who understand the mechanisms of scaling up. Often, people on the lab side don’t necessarily have the experience or the tools to do this. The right people will know to control all the important factors, such as maintaining molecular scale conditions as purification processes are boosted along with the bioreactor process. If the staff members aren’t experienced in manufacturing, they can make incorrect assumptions, and this can present challenges or failures in that scale-up process.

The Timeline from Building a Facility to Rolling Out Product

Typically for these facilities, design can take six months, while construction could take nine months. You then should establish all the systems to operate in a good manufacturing practice (GMP) environment. That can take anywhere from one to two years. You may run into issues in small-scale manufacturing that require you to do further development to support the large-scale processes that may be occurring – maybe a technology isn’t available, or you need to use a different configuration for clarifying your materials. Most often, three years elapse between deciding to build a facility and manufacturing a product that meets all the requirements.

Smaller-scale systems set up for clinical use could be built faster, like when you have existing infrastructure for a GMP operation. If you leverage other companies’ capabilities to validate the equipment before the facility is constructed or hire people to build all the business process systems you’ll use to operate your GMP facility, these can also speed things up. In general, a good estimate is three years in advance of the product launch.

The Typical Process of Manufacturing a Product

Usually, it takes a few weeks to expand the cells to get them to where they need to be. Expansion of the virus may be needed as well to produce enough to support a large-scale batch. Viruses must be applied to the main body of cells in a bioreactor to produce the virus containing the gene of interest. Potentially, the cells may need to be broken open to release the virus, then the addition of an endonuclease to break down the host cell DNA. You may need a depth filter to remove the cellular debris. Then, one or two chromatography steps are needed to purify the viral particles adequately. Finally, there’s a tangential flow filtration to exchange the buffer to a final formulation and a concentration adjustment. The final step is a 0.2 micron filtration to create your bulk drug substance.

Typically, your bioreactor time is in the order of a week. Purification is two to four days, depending upon the complexity of the process and how you spread it out. You’ll need to organize the facility, maximize the turn rate, and support that production as needed for your process.

The Risks Associated with Non-Optimal Purification

Without good purification, immunogenic responses to impurities may occur in the process. But these are vaccine-type processes. There have been many instances of administering vaccines with cellular impurities without causing any severe adverse events to generate the vaccination effect. Good practices will keep microorganisms, endotoxins, and other contaminants out, but it’s unclear whether impurities from the process have a significant impact on the patient. No company wants to risk that. That said, they want to isolate the viral particle and minimize the cellular impurities as best they can.

The Most Common Reasons for Batch Failure

In addition to the manufacturing process, there are also processes to prepare the systems for production. Leaks in single-use disposable systems or failures in the preparation or sanitization could cause batch failures. There are some points of variability in the manufacturing where if you changed the seed stock, the operating conditions you used in the past may no longer be optimal due to the variability in the assay that’s used to measure viral concentration. So you may have additional controls in place – which may not be apparent when you’re working on a small scale – that need to be added to allow you to process robustly and adapt for changes in the viral seed stocks in your process.

Generally, good visibility (real-time viral concentration data of the process) doesn’t occur during processing, so you need to eliminate variability as much as possible when you’re applying the virus to the main bulk of cells. This ensures consistency in production.

How Many Runs Can a Facility Handle in a Year?

It depends on the turn rate of your longest process, which is typically the bioreactor. If the process is seven days, you can run a batch every week. If it’s longer than that, and you can staff the bioreactor to support a seven-day-a-week operation, batches can be run in eight or nine days. An eight-day cycle without weekend shifts could produce product every two weeks. Basically, the bioreactor sets the turn rate of the facility itself.

A small-scale, 50-liter bioreactor may create enough product for years of supply for a particular indication with a small dose. For indications where you’re treating larger organs or the entire body, you could get into a situation where it’s batches per patient, not patients per batch. It comes down to the sizing of the operation according to production needs.

Challenges Facing Gene Therapy Manufacturing

Currently, the industry lacks experienced people, and companies are learning manufacturing alongside the companies they’re providing the service to. It’s not like immunoglobulins, where there’s a 30- to 40-year experience base. For these companies, it comes down to the specific talents they have, and they must pay attention to the details – not doing so shows up in higher failure rates and more operator mistakes while manufacturing products.

The Pros and Cons of Manufacturing In-House versus Outsourcing to a Contract Manufacturing Organization (CMO)

Manufacturing in-house is ideal, but in some cases, it doesn’t make sense. For example, if you have one batch producing many units of product and you can support your entire clinical testing with a few batches, that’s better suited to a CMO. If building your own facility in-house, you need the right people to set it up and operate it in a GMP fashion. That gives you control of the schedule, and if you have indications that need many batches to support your clinical trials, you may find that doing it yourself is a cost advantage. Many companies start off with a CMO, then switch to building their own facilities to support phase 3 and commercial production.

Some companies don’t want to build their own facilities and instead hire a contract manufacturing organization (CMO) to build a dedicated line while still maintaining control of the manufacturing schedule. This way, they don’t necessarily have all the overhead of manufacturing facilities. A thorough financial analysis can help you determine which avenue you can support that can also meet your timeline. Many companies now tell potential customers they’re scheduling production in 18 months, but you’ll need to strategize regarding the path you’ll take and how you’ll manage the risks to your timeline.

Can the Industry Handle All the New Launches?

To meet demand, lots of expansion is needed, both in internal manufacturing and at the CMOs. Timelines can possibly be shortened to get a manufacturing capability established for a reduced cost. For example, instead of building your own water system, buy buffers that are manufactured elsewhere. These types of system validations would delay your facility from becoming operational by having higher purchased item costs and reduced capital costs if you can push them off, you can get up and operational quicker. Companies can also save time by leveraging CMOs where it makes sense, such as in-filling lines where there may not be a constraint in capacity.

The Biggest Challenges that Need to Be Resolved in the Next Five Years

The largest challenge is establishing robust processes to manufacture products, but there’s a lack of in-process measurements that provides a good picture of what’s going on. As I monitor a manufacturing process, I look at all the data available from the process. Is it telling me the same story or not in terms of that process? Are you monitoring the media you’re using to ensure you’ve got the right nutrient levels? Your media may be producing lots of cells, but maybe those cells are consuming a key item in your viral production. There’s a lot of complexities. If we can put in the necessary controls for that robust production so that it’s repeatable, that will solve the biggest challenge in the consistent manufacturing of products.

---

About Lance Weed

Lance Weed was most recently Vice President of Operations for uniQure where he designed, built, and established operations for the largest gene therapy manufacturing facility using 100% single-use disposable processing systems. Prior to his role at uniQure, Lance worked for BioVex as VP of Operations, where he designed, built, and established operations for an oncolytic virus manufacturing facility which was purchased by Amgen in 2011 for $1bn.

---

This article is adapted from the GLG teleconference, Gene Therapy Outlook – A Manufacturer’s Perspective. If you would like access to this webcast, or to speak with Lance Weed or any of our more than 700,000 experts, contact us.