The past year revealed both successes and setbacks for viral-vector gene therapies. The rapid development and large-scale rollout of multiple adenovirus-vector vaccines represented an unprecedented achievement that is poised to help mitigate the devastating impact of the COVID-19 pandemic. During the same period, multiple high-profile gene-therapy assets encountered challenges, with clinical trials paused because of safety concerns or failing to meet efficacy targets.
These successes and setbacks are emblematic of the current state of viral-vector gene therapy: a technology with considerable promise but with a set of challenges still ahead. As more and more gene therapies have reached the clinic, it has become clear that multiple technological challenges must still be overcome to unlock the full potential of viral-vector gene therapy.
Rising to meet these challenges, biotech and pharmaceutical companies are testing a multitude of technological advances and innovative strategies that address all aspects of viral-vector gene-therapy development. For companies prepared to keep abreast of the rapid pace of change, these innovations offer a path for ushering in the next generation of viral-vector gene therapies.
Companies are testing a multitude of technological advances and innovative strategies that address all aspects of viral-vector gene-therapy development.
The state of viral-vector gene therapy
Viral-vector gene therapies use modified viruses as drug-delivery vehicles to introduce specific DNA sequences—encoding genes, regulatory RNAs (for example, small interfering RNAs [siRNAs]), or other therapeutic substrates—into cells. The technology has long drawn interest for its potential advantages over traditional modalities. Many types of therapeutic agents (for example, enzymes, antibodies, and siRNAs) can be encoded in DNA sequences that can be rapidly designed and synthesized once a target is identified.
Viruses serve as powerful delivery vehicles for these sequences because of their ability to enter cells efficiently and potentially gain access to hard-to-reach, highly specific cells. In combining these features, viral-vector gene therapies can be used to modify gene expression in a programmable way, offering the flexibility to potentially treat a wide spectrum of diseases—including rare monogenic diseases by gene replacement and broad-population diseases by controlling gene expression—and help disease prevention by immunization.
Nearly all gene therapies currently available use one of three vector types: adeno-associated-virus (AAV) vectors, adenovirus vectors, or lentivirus vectors (Exhibit 1). AAV and adenovirus vectors are typically used in gene therapies that are directly administered to patients by infusion or local administration (in vivo), with AAV being the most popular vector for areas outside of oncology and vaccines. Lentivirus vectors are typically used for ex vivo therapies, in which cells harvested from a patient are modified in the lab before retransplantation. This article primarily focuses on in vivo gene therapies; however, many of the challenges and advances discussed are applicable across both routes of administration.
Excitement around viral-vector gene therapies is evident. While only four in vivo viral-vector gene therapies are currently on the market, more than 100 gene-therapy assets are in clinical trials as of late 2020, with a far greater number in preclinical development.
Many of these assets have emerged from the steady stream of small- and midsize biotech companies and academic labs supported by continued, high levels of venture-capital funding. Large pharma companies have increasingly focused on the potential of viral vectors, with seven biotech-company acquisitions valued near or above $1 billion in the past two years alone (Exhibit 2).1 Adenoviruses are being proven as a vaccine platform, with approvals for Ebola vaccines and groundbreaking COVID-19 vaccines over the past year.2
While the high list-price of some gene therapies was once seen as a near insurmountable challenge to commercialization, innovative reimbursement strategies have shown that successful launches are possible, with ZOLGENSMA (treating more than 600 infants with spinal muscular atrophy3 in its first ten months on the market) beating analyst expectations.4 Worldwide sales of viral-vector gene therapies are forecast to grow at a rate of more than 50 percent year-on-year for the next five years (excluding the potential impact of COVID-19 vaccines), affecting the lives of tens of thousands of patients.
However, while there is significant momentum, there have also been multiple recent setbacks.5 Many of these relate to challenges previously outlined by McKinsey in its perspective on the future of gene therapy (including efficacy, durability, and manufacturing). As these therapies have sought to expand beyond the ultrarare indications they originally targeted, three technological challenges have emerged as recurrent obstacles. For viral-vector gene therapies to reach their true transformative potential—much like monoclonal-antibody technology 20 years ago—this set of technological challenges must be overcome.
Challenges to realizing the potential of viral-vector gene therapies
The current generation of viral-vector gene therapies represents the culmination of decades of biological and clinical research. As more patients have received these therapies, it has become clear that three fundamental challenges will restrict the applicability of viral vectors: getting past the immune system, lowering the dose, and controlling transgene expression. Ongoing work to address these challenges is generating technological innovations that have the potential to leapfrog current therapies and unlock the potential of viral vectors.
Getting past the immune system
The success of any viral-vector gene therapy depends on its ability to get past multiple lines of defense deployed by the human immune system. Viral capsids, viral-vector DNA, and even the transgene products themselves may be recognized as foreign, providing multiple opportunities for the immune system to clear the gene therapy from the body.
Immunity against viral capsids can limit the efficacy of a gene therapy. Because most viral-vector gene therapies today use vectors derived from harmless viruses circulating in humans, many patients (up to 60 percent) may have preexisting immunity from past exposure.6 CanSinoBIO, for example, reported reduced efficacy of its COVID-19 vaccine in individuals with preexisting antibodies to the adenovirus-5 (Ad5) vector it chose for drug delivery.7
Although this effect depends on the vector serotype used, and the clinical impact is still unclear,8 many clinical-trial sponsors conservatively exclude patients from their studies if they have antibodies to the vector in question. This can come at the cost of making most patients ineligible for therapy. Acquired immunity to viral vectors poses additional challenges for viral-vector gene therapy in the long term. Patients treated with a gene therapy today may not be able to receive a second gene therapy in the future if the same viral vector is used in both contexts.
In addition, viral capsids and viral-vector DNA can actively provoke an immune response from the body. For viral-vector vaccines, this immunogenicity can be beneficial, as it reduces the need for adjuvants and increases efficacy. However, for other viral-vector gene therapies, immunogenicity can reduce efficacy, increasing the chance that the gene therapy is detected and eliminated by the immune system. Indeed, some have speculated that immunogenic vector DNA sequences are behind the limited durability of some recent gene therapies, leading to their abandonment.9 More concerningly, immunogenicity can lead to safety concerns during therapeutic use, as high levels of viral capsids can cause severe immune reactions at the time of injection.
Unraveling the immune system’s intertwined responses to viral-vector gene therapies remains difficult. Animal models do not recapitulate all relevant aspects of the human immune system (as immune systems behave quite differently among species). While human clinical trials offer a valuable source of insight, many gene-therapy trials are too small to confidently isolate the parameters associated with a drug’s success or failure.
Lowering the dose
Current viral-vector gene therapies require the administration of large numbers of viral particles to patients, particularly for therapies aimed at treating systemic diseases. For example, recent gene therapies for Duchenne muscular dystrophy (DMD) that aim to correct mutations in muscle cells throughout the body have delivered up to approximately 10^16 (ten-thousand trillion) viral particles in a single dose (for example, a dose of 3 × 10^14 vector genomes [vg] per kilogram [kg], assuming a 30-kg child),10 which is multiple times the number of cells in the human body.11 For systemic diseases, the need to individually target and repair many cells in the body partly explains why such large doses are administered. Another explanation is the limited cell-type specificity of current viral vectors: large numbers of viral particles must be delivered to ensure that an adequate number reach clinically relevant cells.
The large doses used in current gene therapies pose two challenges. First, large doses are difficult and expensive to manufacture. Today, a typical manufacturing run of an AAV-vector therapy using high-yield cell lines and large-capacity bioreactors might only produce approximately ten doses of a systemic gene therapy from a single batch at a cost of nearly $100,000 per dose (assuming approximately 1 × 10^17 vg per batch).12 Although these costs will gradually decrease as gene therapies begin to reach clinical and commercial scales, any technological advance that reduces the required dose would bring immediate benefit, as a tenfold reduction in dose might also bring about a tenfold reduction in costs.
Second, and even more critically, administering large doses of virus has been linked to adverse safety outcomes.13 Although investigations of four deaths in clinical trials of AAV-vector therapies in 2020 are ongoing, three deaths occurred in high-dose cohorts. Clinical-trial protocols have subsequently been revised to limit viral dosage, reflecting the tremendous importance of this issue.14
Controlling transgene expression
Once a viral vector successfully delivers its therapeutic gene to the cells in question, the efficacy of the gene therapy depends on the quality of transgene expression. Specifically, the transgene must be expressed at the appropriate level (neither too low nor too high), in the appropriate cells, and for the appropriate duration to mediate the desired clinical effect. For therapeutic uses (in contrast to use for vaccines), the transgene may need to be expressed permanently if the gene therapy is to serve as a one-time cure and represent an appealing alternative for patients over current standards of care requiring repeated dosing (which may not be possible because of the challenges previously laid out). Regulators have required multiple years of follow-up data showing that gene expression is maintained. Indeed, some drugs have been abandoned when expression waned after 12 months.
To maximize chances of success, early viral-vector gene therapies have opted to include regulatory elements (DNA sequences such as promoters and enhancers that control how genes are expressed) that have been selected to drive high levels of transgene expression in all cell types. However, this approach may have significant drawbacks, particularly as gene therapies move beyond gene replacement for monogenic rare diseases. Overexpression of the transgene or its expression in the wrong cells may contribute to inflammation and other toxicities (as was observed in recent studies of nonhuman primates).15 Moreover, current gene therapies, once administered, cannot be controlled or turned off by clinicians should the need ever arise.
Innovative solutions that address gene-therapy challenges from many angles
To tackle the challenges facing gene therapy, academic labs, start-ups, and established companies are generating myriad innovative solutions (Exhibit 3). Each focuses on a specific component of a gene-therapy product (for example, the viral capsid) or part of the development process (such as manufacturing). However, these innovations often address multiple core challenges, outlining multiple paths to realizing the promise of viral-vector gene therapy.
We have identified five key trends to watch.
1. Improved capsids
The viral capsid is a critical component of viral-vector gene therapy. It determines which cells are targeted, the efficiency of cell entry, and the probability that the gene therapy is detected and eliminated by the immune system. In addition, the capsid is largely responsible for the stability of the viral vector during the manufacturing process and can affect storage and distribution requirements.16
The capsids most widely used today, including those used in on-market products, are derived from naturally occurring viruses. They have suboptimal properties, including little cell-type specificity, moderate efficiency of cell entry, and relatively high levels of preexisting immunity in humans. To address the problem of preexisting immunity, many assets use capsids from viruses found in other species. For example, the AAV8 and AAVrh74 capsids used in multiple AAV-vector gene therapies are derived from AAV serotypes isolated from macaques, and some of the COVID-19 vaccines that have been developed have used adenovirus serotypes from chimpanzees and gorillas. While this approach may limit the challenges of preexisting immunity, it largely doesn’t address specificity or efficiency (particularly as these viruses have evolved to infect nonhuman species).
Increasingly, drug developers are turning to capsids that have been engineered in the lab and can be selected to overcome the challenges mentioned previously (Exhibit 4). These engineered capsids are identified through large-scale screening efforts in which millions of variant capsids are screened for the desired properties and iteratively refined. Capsid-engineering platforms—many of which have been spun out of academic labs to form companies—achieve these ends by leveraging advanced technologies, such as cryo-electron microscopy (cryo-EM) and artificial intelligence.
Improving capsid properties could bring multiple immediate benefits. For example, a twofold increase in a capsid’s cell-type specificity could enable a twofold decrease in the overall viral dose required, thereby improving safety and cost. It’s still too early to determine the true impact of capsid engineering, as most engineered capsids are still in preclinical development. However, companies’ early reports suggest that capsids with five- to tenfold improvements in multiple attributes may be entering the clinic soon.
2. Improved vectors
Like the capsid, the DNA sequence of the viral vector itself affects multiple aspects of a gene therapy’s performance, but engineering the vector can often be considerably easier, cheaper, and quicker. Accordingly, vector engineering is becoming a growing focus of gene-therapy R&D. Vector engineering is often easier with adenovirus and lentivirus vectors than with AAV vectors because of AAV’s inability to package large pieces of DNA. However, innovative vector elements are beginning to appear in AAV-vector designs as well.
Vector engineering broadly has two aims: reducing the immunogenicity of the viral vector and improving transgene expression. One strategy to achieve both aims is codon optimization, in which variations in the vector sequence are explored to eliminate immunogenic sequence motifs while optimizing the transgene for robust expression. Subtle changes in vector sequence achieved through codon optimization can have large effects, such as increasing expression levels and possibly extending the duration of expression for multiple years.17
Vector engineering broadly has two aims: reducing the immunogenicity of the viral vector and improving transgene expression.
Transgene expression can be further programmed by engineering regulatory elements into the vector sequence. Some regulatory elements turn on transgene expression only in certain cell types or tissues—ideally, the disease-causing cells—preventing potentially toxic expression in other contexts. Such cell-type- or tissue-specific regulatory elements (for example, promoters and enhancers) have become relatively common in viral-vector gene therapies. For an additional layer of control, some viral-vector gene therapies are also incorporating regulatory elements, such as microRNA-target sites, that reduce expression in specified cells—for example, in cells that promote an immune response.
Finally, a more distant and challenging goal is to engineer vectors that are inducible, where transgene expression can be controlled using an additional signal, such as an orally administered small-molecule drug. This could allow clinicians to turn on, turn off, or otherwise adjust a gene therapy after it is administered, delivering a personalized course of treatment.
3. New types of cargo
The cargo delivered by a viral-vector gene therapy is typically a working copy of a gene that is used to replace the patient’s disease-causing copy of that same gene. However, any therapeutic agent that can be encoded in DNA can theoretically be delivered by a viral vector. Researchers and drug developers are increasingly leveraging this flexibility to deliver other types of molecules with therapeutic value—alone or sometimes in combination—including regulatory RNAs (for example, short hairpin RNAs [shRNAs]), vectorized antibodies, and substrates for gene editing.
Gene editing is an intriguing potential solution for achieving long-lasting, physiologically appropriate gene expression. For patients with diseases caused by certain types of mutations, restoring the function and expression of the patient’s own copy of the gene through gene editing may be simpler (and more permanent) than attempting to engineer and deliver a replacement.
4. Improved manufacturing processes
Early gene-therapy-manufacturing processes originated in academic labs and were focused on small, research-scale batches. These processes were not optimized for moderate- or large-scale production or for the delivery of systemic therapy. As gene therapies start to expand outside the treatment of ultrarare diseases, one of the many challenges being addressed is the presence of empty capsids created during the manufacturing process. These empty capsids, which have no active cargo, can create the requirement for higher doses and, accordingly, stimulate stronger immune responses.
Two approaches are being developed to reduce the ratio of empty-to-full capsids in manufacturing: developing improved methods to separate the empty from full capsids based on specific properties (for example, charge and molecular weight) and engineering cell lines that package full capsids more efficiently. By reducing the empty-to-full ratio, these advances reduce manufacturing costs, reduce immune responses, and improve the safety of gene therapy. Indeed, regulators have used reducing the empty-to-full capsid ratio as part of the rationale for lifting clinical holds on gene-therapy products with previous safety issues.18
5. Improved pretreatment and conditioning regimens
Beyond engineering the capsid and vector, a separate approach for reducing the immune system’s detection of viral-vector gene therapies involves coadministering the therapy with an immunosuppressive agent. Multiple such conditioning regimens are currently being tested to reduce the impact of neutralizing antibodies on the efficacy of the treatment, both of preexisting antibodies and newly generated antibodies that could prevent future redosing. Nearly all current viral-vector gene therapies use steroids to help manage the potential immune response to the viral vector; however, the type, dosage, and timing of the steroid treatment varies widely.
Some clinical trials are experimenting with more targeted immune suppression, such as the use of rituximab to reduce the creation of memory B cells.19 An even greater assortment of approaches is being tested in animal models to directly reduce the presence of neutralizing antibodies. These include the use of enzymes cleaving to immunoglobulin G (IgG), plasmapheresis to remove the neutralizing antibodies specific to the gene therapy, and even CRISPR-based repression of neutralizing-antibody creation.20 These approaches could expand the pool of eligible patients to include those with preexisting immunity. Moreover, these approaches could enable a patient to receive multiple doses of the same therapy or of different therapies using the same vector backbone.
The road ahead
Viral-vector gene therapies find themselves at another inflection point. Early successes in the treatment of rare diseases and vaccines have proven the potential of this modality, while the challenges to gaining widespread adoption—the way that monoclonal antibodies have over the past 20 years—have only become clearer. Nevertheless, the wealth of innovative solutions being explored across academia, biotech, pharma, and contract development and manufacturing organizations demonstrate that viral-vector gene therapies are here to stay.
As described previously, different solutions are emerging to address each of the core challenges. The diversity of these approaches and the complexities of gene therapy mean that no single approach is likely to “win.” That situation will enable a rapid innovation cycle in which gene therapies are constantly being improved upon, which will offer new opportunities to leapfrog existing products. Even as AAV-vector-based delivery is becoming the leading technology, some prominent limitations combined with the rapid pace of innovation leave the door open for other delivery technologies to emerge.
Owners of viral-vector platforms will need to consistently look to the next set of innovations beyond their current platforms and assets. That could include investing directly to help overcome the broader challenges or buying or licensing critical technology to upgrade their platforms. Indeed, multiple new biotech companies have launched to solve one or more of the challenges outlined in this article as a service to developers of gene therapies. Staying abreast of these developments will require fastidious monitoring of scientific and technological progress on all fronts. However, since it is difficult at this early stage to place bets across all potential solutions and innovators, gene-therapy leaders will need to make their investments judiciously.
In the short to medium term—while technological challenges limit the scope of gene therapies to curative treatments for rare diseases—fast followers may find it difficult to be successful, even with improved technologies, as first entrants rapidly address prevalent populations. Gene-therapy leaders will therefore need to strike a careful balance by accelerating programs today while retaining the flexibility to adopt innovative technologies that unlock treatments for broader-population diseases and the full promise of viral-vector gene therapies in the long term.