Adeno-associated virus (AAV) vectors may be the best option to deliver genetic cargo to cells in the body. But they are far from perfect. They require large doses, hold small amounts of cargo, and can be recognized by the immune system—a safety concern. Moreover, they perform differently in animal models such as mice and nonhuman primates than in humans. These complications demand ongoing innovation and the development of new strategies.
Recently, the discrepancy between the AAV-associated immune response in animals and that in humans stung LogicBio Therapeutics. The well-established company had to suspend a clinical trial when two participants experienced adverse events. Just starting out, the company Akouos is hoping to avoid some of the pitfalls by working in an area of the body, the inner ear, where the company’s AAV technology has shown promise.
Learning along the way
Mark A. Kay, MD, PhD, professor of genetics at Stanford University, started working in gene therapy in 1990. Back then, he recalls, people would give five-year estimates for when some diseases would be cured.
Kay was part of the team that filed an IND application in 2000 that led to the first trial in which AAV2 was systemically administered to humans. This work was done with gene therapy pioneer Katherine High, MD, when she was affiliated with the University of Pennsylvania. (Today, she is a visiting professor at the Rockefeller University and the president of therapeutics at AskBio.)
Over the years, Kay has learned that AAV dosages and transduction efficiencies in animal studies don’t necessarily correlate to those in humans. Simply put, predictions made from animal studies don’t correlate to humans.
For example, during that early trial 22 years ago, the team saw no toxicity when the AAV vector was given to dogs and rodents. But there was liver enzyme elevation in humans. Also, in humans, but not in animals, there was a T-cell response against the capsids that had been taken up by cells and degraded.
This experience led Kay to find a better model for AAV vector selection. This model, a mouse that was developed in the laboratory of Markus Grompe, MD, director of the Papé Family Pediatric Research Institute at Oregon Health & Science University, incorporates primary human hepatocytes. Using this model, Kay’s group became the first to use “humanized” mice to compare the transduction efficiencies of different AAV vectors.
Kay’s group was also the first to use an AAV vector development process in which capsid variants were selected from a capsid library. In this process, scientists disassemble nature’s capsids, shuffle the resulting capsid pieces, and then randomly assemble new capsids. Ultimately, the capsids become part of capsid libraries, which may contain millions of capsid variants.
The capsids are injected into the humanized mice, and then the capsids that succeed in transducing human hepatocytes are recovered. One of the most successful capsids found by Kay’s group is the AAV-LK03 capsid. It works 10 times better than either AAV2 or AAV8. It was used in the LogicBio trial as well as in a trial run by Spark Therapeutics. Spark called the capsid Spark200 and used it to deliver SPK-8011, a gene therapy for hemophilia A.
Kay screens for AAV vectors that have different properties. For example, he screens for AAV vectors that can package a larger amount of DNA. Screening for AAV vectors that don’t elicit an immune response is more difficult, he says. Unless you know more of the molecular details about what stimulates those responses, it is more difficult to screen for them.
A first: In vivo genome editing in children
Together with two postdoctoral researchers, Kay spun his laboratory’s work out to form LogicBio in 2014. In 2016, LogicBio welcomed its new president and CEO, Fred Chereau, an executive with extensive experience building biotechnology companies.
In October 2021, LogicBio announced clinical trial results demonstrating the first-ever use of nuclease-free genome editing technology in children. This technology was also developed in Kay’s laboratory.
LogicBio focuses on the development of gene therapy solutions for pediatric patients with rare diseases. The first drug in the company’s pipeline (LB-001) is for methylmalonic acidemia (MMA), a life-threatening condition affecting approximately 1 in 50,000 newborns in the United States.
Early data from the Phase I/II trial looked promising. Based on the safety data from first two patients, the independent Data Safety Monitoring Board recommended continuation of the trial, enabling the enrollment of younger children and the use of higher doses.
But on February 2, the company hosted a conference call to reveal that the FDA had placed the MMA trial on clinical hold. At that time, four patients had been dosed. The first two patients, the older patients, were fine. But the third and fourth patients experienced thrombotic microangiopathy (TMA), an adverse event previously reported in other AAV gene therapies.
Both patients are doing well now and are back home, Chereau says. At LogicBio, investigators are working to implement changes to get the trial back up and running. Although the TMA phenomenon is not that well understood, notes Chereau, it is known that complement activation is involved. Although it is now standard to give steroids prophylactically with a high intravenous dose of AAV vector–delivered therapeutic, more is needed to prophylactically mitigate the response. To do that, LogicBio’s investigators can amend the protocol to add a product (a C5 inhibitor) that will tamp down the complement activation.
Chereau says that gene editing and gene therapy are still in their early days, and that companies in the field are still learning. It took time for the field to learn how to introduce steroids. Now, people in the field know that steroids are good but may not be sufficient. So, these people are considering other changes to their protocols. For example, in an IND application for a DMD program submitted by Regenxbio, a protocol is described in which an AAV8-based gene therapy incorporates a complement activation mitigation strategy.
Some researchers are focusing on other workarounds. Some make “stealth AAVs” or try to engineer out epitopes that flag an AAV to the immune system. It’s unclear if these will work, notes Kay. He favors the strategy of finding AAVs that can be used at lower doses.
The gene therapy field is being informed as development programs proceed, and it is reacting to the good and the bad. Chereau notes that LogicBio has been contacted by other gene therapy companies that are thinking about their own IND applications. And LogicBio has found that its problems are shared by other companies, and that these companies are willing to share information. When it comes to the safety of the patients, the industry is willing to collaborate. Chereau declares, “We are all learning from each other.”
Gene therapies for inner ear conditions
An ear for music may accompany a mind for science. Just look at the Longwood Symphony Orchestra (LSO), a Boston group made up of scientists, physicians, and other healthcare professionals from the Longwood medical area—home to Harvard Medical School and several hospitals.
The LSO’s scientist-musicians aren’t unique. Another scientist-musician, one who is also active in the Boston area, is Manny Simons, PhD, CEO of Akouos. Simons studied music when he was an undergraduate at Harvard College. But his interest in neuroscience started competing for his attention.
Indeed, the balance shifted to neuroscience when Simons chose to study biomedical engineering at the Massachusetts Institute of Technology (MIT). In the laboratory of Robert S. Langer, ScD, Simons focused on inner ear drug delivery. In 2008, he earned his doctorate, and in 2014, he joined Voyager Therapeutics, an AAV gene therapy company. But he never lost his passion for music. For Simons, this passion strengthened his appreciation for hearing.
In 2016, Simons co-founded Akouos, which is focused on inner ear conditions. Ever since, Simons has been able to combine an interest in science and a commitment to developing treatments to restore hearing. His company has two leading programs. By the end of the year, both may figure in IND applications. If they go forward, these two programs will be the first two inner ear AAV programs to enter clinical development.
The first program, AK-OTOF, is an AAV gene therapy for the restoration of hearing in individuals with a genetic sensorineural hearing loss due to mutations in the otoferlin gene. The otoferlin gene is expressed in the sensory cells (hair cells) that are found in the inner ear and that convert sound-driven fluid waves into neural signals. The second program, AK-antiVEGF, is an AAV gene therapy for the treatment of vestibular schwannoma. AK-antiVEGF is designed to enable cells in the inner ear to secrete an antitumor protein.
To date, there has never been a single drug or biologic approved for direct inner ear administration. The inner ear is not easy to access. It is closer to the brain than the middle ear, and the sensory cells are encased in a fluid-filled, bony structure.
AAV gene therapies are “ideally suited to the inner ear,” Simons asserts. The AAV vectors can be delivered directly without exposure to the rest of the body, allowing a high ratio of vector to contact the target cells. Also, the hair cells don’t divide, which means the expression of functional genes may be sustained over the long term. (Because AAV vectors deliver the gene as an episome—DNA that doesn’t integrate into the cell’s genome—division means dilution.)
In addition, AAV vectors are relatively well tolerated in the inner ear, where immunity is reduced. In Akouos’s tests with nonhuman primates, AAV vectors did not provoke an immune response, even when they were administered in high doses. Nonetheless, Akouos uses AAV vector doses lower than those used in traditional gene therapy. Normally, the dose of AAV vector delivered to the bloodstream is 1014 or 1015 viral particles per kilogram. The doses Akouos uses are three or four orders of magnitude less than that. Collectively, these factors remove a lot of worry from Akouos’s work.
A fortuitous confluence of factors
When Akouos was started, Simons and the company’s other co-founders, Michael McKenna, MD, and Bill Sewell, PhD, were focused on the surgical delivery approach needed to deliver vectors to the inner ear. But they didn’t have a vector. Indeed, at the time, it was not known if there were capsids with a tropism for the inner ear.
Before long, Akouos enjoyed a stroke of luck. It came from the laboratory of Luk Vandenberghe, PhD, director of the Grousbeck Genn Therapy Center at the Massachusetts Eye and Ear. Vandenberghe had built an ancestral AAV capsid library by using the capsid sequence of all known naturally occurring AAV capsids and computationally predicting their ancestors. Then, he developed a synthetic AAV vector that could enable safe and efficient gene transfer to the mammalian inner ear (Landegger et al. Nat. Biotechnol. 2017; 35(3): 280–284).
Vandenberghe showed that an AAV vector, Anc80, has a high transduction efficiency for the hair cells in the inner ear. Akouos licensed the entire library (38,000 distinct capsids) for all inner ear uses. Marrying together Vandenberghe’s Anc80 with the surgical delivery method for the inner ear meant that the company was on its way. Simons said that these factors “came together nicely to open up a new field of medicine.”
So, does anything worry Simons about the trial? It is the first of its kind, he admits, and it is enrolling children. And the company won’t know, until the trial starts, whether the initial dose will have an effect—or whether the children will see benefits.
Gene therapy hurdles are numerous and difficult to clear. Even with the complications of the immune system diminished, as is the case for Akouos’s work in the inner ear, the questions of efficacy loom large.
Recalling his start in the field, Kay says that even modest advances were celebrated: “If we could turn one cell positive in a mouse, we would jump up and down.” Today, he finds inspiration in the development of monoclonal antibodies and bone marrow transplantation. These modalities, which took decades to establish, show how patience can be rewarded.
Many different types of approaches are helpful. As Kay notes, “We don’t know which platform is going to be superior.” There are right and wrong ways to do things, he adds. But there is not one right way to go forward.