Immuno-Oncology—The Translational Runway for Gene Therapy: Gene Therapeutics to Address Multiple Immune Targets

Introduction

Engaging the immune system is regarded as the most promising strategy to combat cancer, a view that is supported by the breakthrough success of checkpoint inhibitors. ¹ However, it has also become clear that for effective immunotherapy, addressing one target at a time is not sufficient. The hurdles for the translation of cancer immunotherapy are high, as described in an international effort in 2011. ²

This poses a problem for the biopharmaceutical industry as well as for regulators, because both have long since followed the “single target–single drug” paradigm. They are now experiencing a historical challenge to their traditional concepts. Combination treatments are a potential quick fix for this problem, but drug candidates and drugs for use in such combinations are rare, and the individual elements have been neither developed nor approved for use in combination, meaning that this strategy is merely a provisional one. In addition, it is much more difficult to find predictive animal models to assess preliminary safety and efficacy in humans, posing new challenges in the preclinical stage.

Obtaining regulatory approval for combination therapy is therefore largely unknown territory. Safety, additive effects, effectiveness, and dose–response need more thorough investigation than when developing individual drugs. In June 2013, the Food and Drug Administration (FDA) published guidance for industry addressing these problems. ³ Nevertheless, combination therapies are being explored in non-cancer indications, for example in macular degeneration and Alzheimer's. ⁴

There may be an elegant and easier way to overcome this complex situation, however, from a field that—without generating any hype—has already developed into a cornerstone of today's most advanced immuno-oncology therapy strategies: gene therapy.

For more than two decades, the field has failed to live up to the high hopes for the therapy of inherited diseases. Gene therapy was characterized by clinical setbacks rather than breakthroughs. But the efforts of these two decades have not been in vain, as they have resulted in a wealth of experience with vectors and delivery, as well as with the organization, regulation, and stabilization of genetic content. ⁵ The lessons learned have now led to rich innovation in cancer therapy and may well lead to breakthrough treatments against tumors. At present, there are already about 400 studies listed in the clinicaltrials.gov database exploring gene therapy as a cancer treatment. They are employing a broad number of immuno-oncology treatments involving gene therapy, and regulators classify most of them, including cell therapy products, as gene therapy already.

These approaches demonstrate the progress that has been made with the design of various gene therapy vectors. Genetic vectors can not only be used extracorporally to reprogram cells, but can also be delivered precisely to target tissues and cells and expressed in the desired fashion to reach a therapeutically effective concentration for a desired period of time.

However, while the above-mentioned approaches are all based on genetically engineered vectors, there are remarkable differences, demonstrating the versatility of today's gene therapy technology. For one, the vectors range from engineered live viruses to non-integrating, non-replicating viruses to naked or packaged nucleic acids. Second, the vectors are used for very different approaches. Some still follow the single target–single drug paradigm and deliver a therapeutic agent such as interferon (IFN) alpha 2b or proapoptotic human Fas-chimera directed against a specific metabolic or other cellular property. Other approaches follow the combination therapy paradigm and are using genetic vectors to combine various therapeutic active ingredients elegantly in one drug.

The simplest approach is combining the genetic information for several antigens in one vector to design an anticancer vaccine directing the immune system against different tumor targets. These antigens still trigger a common mechanism. While preliminary data are encouraging, it is still an open question whether these vaccines, even when they are based on neo-antigens, will be sufficient for a strong and lasting efficacy.

While the speed with which these diverse technologies have been introduced in the market is remarkable, there is still a host of problems with these gene therapy approaches in cancer. The common problem in most vaccination approaches is poor efficacy because the immune system does not produce enough T cells for a robust, lasting reaction and because tumors develop a wealth of evasion mechanisms that deactivate the immune reaction triggered in the first place.

The most sophisticated approaches therefore employ gene therapy to ensure the production of diverse therapeutic molecules at the target site. That way, a single drug is able to address different targets to trigger different mechanisms at once so that several tumor evasion strategies are addressed at once.

Extracorporal Applications

The genetic modification of immune cells outside the body avoids problems arising from off-target effects observed in systemic gene therapy applications. However, many approaches demonstrated poor efficacy in clinical trials or posed serious safety problems.

DC vaccines have shown considerable success in treating cancer, and with Provenge sipuleucel-t, a first DC vaccine has been approved for the treatment of advanced prostate cancer. However, preparation of this autologous treatment is tedious, time-consuming, and very costly. To improve targeting, enhance immunogenicity, and reduce susceptibility to the immunosuppressive tumor microenvironment, DC vaccines have been engineered with a variety of physical methods or different viral vectors. Physical methods have mostly resulted in almost undetectable transgene expression, while viral vectors have proven much more effective, with adenoviral and retroviral vectors demonstrating the highest effectivity. However, while preclinical results were often promising, clinical responses so far have been poor. ^15,16

Early in 2017, a Phase III clinical study of the personalized autologous DC immunotherapy rocapuldencel-T in patients with metastatic renal cell carcinoma was ended following an interim analysis by the study's independent monitoring committee. The committee concluded that the study was unlikely to demonstrate a statistically significant improvement in overall survival in the treatment arm. ¹⁷ In the study, the personalized vaccine was combined with Sutent sunitinib, an oral, small-molecule, multitargeted receptor tyrosine kinase (RTK) inhibitor of cellular signaling. Rocapuldencel-T is produced from isolated patient tumor mRNA, which is then transferred into patient DCs by electroporation. Recent DC vaccination technology for acute myeloid leukemia (AML) promises considerable improvements in manufacturing and efficacy. ¹⁸

NK cells have the ability to recognize and kill tumor cells without requiring prior antigen exposure, and they are therefore seen as promising agents for cell-based cancer therapies. By genetically modifying NK cells to produce cytokines such as interleukin (IL)-2 or IL-15, their survival capacity and proliferation increase, and their activation and antitumor activity in vivo are enhanced. Furthermore, they can also be modified to enhance the specificity for the target cells. However, no clinical trial has clearly demonstrated a significant benefit in tumor patients. ⁶ Early in 2017, the first ever clinical evaluation of a genetically engineered, allogenic, off-the-shelf NK cell for the treatment of patients with cancer began, with results expected in 2018. The NK cells have been engineered to produce endogenous and intracellularly retained IL-2 and to express CD16, the high-affinity (158V) Fc gamma receptor (FcγRIIIa/CD16a). ⁶

Cancer immunotherapies based on genetically engineered NK cells and DCs do not pose serious side effects or safety concerns, but there is still a discrepancy between the effect on the immune system and clinical efficacy. ⁶

CAR-T are T cells with engineered receptors, providing them with the specificity of a monoclonal antibody. Typically, retroviral vectors are used to transfer the coding sequences, which usually are derived from different sources. In cancer, the T-cell antigen receptor is engineered to become specific for malignant cells.

The first generation of engineered CAR-T cells showed little evidence of antitumor activity in clinical trials. Second- and third-generation CAR-T cells with added sequences to enhance activation and proliferation and to augment potency showed better antitumor efficacy in clinical trials but led to severe and sometimes life-threatening adverse events. For example, in one trial, 5/68 patients receiving the therapy died during the study, three of them from developing cerebral edema and two because of similar neurotoxicities. As a result, the FDA established a pilot database to evaluate CAR T-cell safety and identified specific safety-related factors associated with the conduct of clinical trials.

The most serious short-term side effect associated with CAR T-cell therapy is cytokine release syndrome (CRS), a reversible yet potentially life-threatening condition mediated by the release of IL-6, tumor necrosis factor alpha, and IFN-γ following immune cell activation. In addition, CAR T-cell therapies specifically targeting the B-cell-specific marker CD19 for the treatment of blood cancers may result in B-cell aplasia, and some patients receiving non-CD19-targeted CARs have experienced serious toxicities because the same target is expressed on healthy as well as tumor cells. In its recent review of Novartis's CAR-T therapy Kymriah tisagenlecleucel for acute lymphoblastic leukemia (ALL), the FDA raised serious safety questions: short-term and potentially lethal dangers, from a high rate of CRS to the neurological conditions that CRS can trigger to potential long-term safety concerns, including the potential for the generation of replication-competent retrovirus (RCR) and the potential for insertional mutagenesis to cause new malignancies. ¹⁹ In August 2017, the FDA approved Novartis's CAR-T therapy Kymriah. However, the product will carry a boxed warning for CRS and for severe or life-threatening neurological symptoms, among others, and is available only through a restricted program under a Risk Evaluation and Mitigation Strategy (REMS) called the KYMRIAH REMS. Remarkably, it is the first gene therapy ever approved in the United States. ¹⁹ While these concerns may not hamper the development of CAR-T therapies, they are likely to slow it down substantially.

The CRISPR-associated 9 (CRISPR-Cas9) system has made genome editing very convenient and efficient. It has allowed for the rapid introduction of genetic modifications in all sorts of cell lines, organs, and animals. It adds gene editing to whole genome screening and can be used for loss and gain of function modifications. The technology not only eases the design of cancer animal models, but has also been used in medical research for cancer genetics, the large-scale search for cancer driver genes, and functional studies. ⁷ Genome editing using transcription activator-like effector nucleases (TALEN) technology was successfully applied for CAR-T cell optimization. ²⁰

In 2016, it was first used in clinical studies with patient cells. Chinese researchers edited the genome of the lymphocytes of cancer patients to disable the gene encoding for PD-1, so that tumor cells can no longer inhibit the immune response via ligands binding to PD-1 protein receptor. ²¹ In the same year, researchers in the United States were given green light for a Phase I clinical study in which a cancer patient will be treated with autologous T cells modified extracorporally by three edits made with CRISPR-cas9: one edit to insert the gene encoding for the NY-ESO-1-binding CAR to detect and eliminate cancer cells, a second to remove the endogenous T-cell receptor that could interfere with this process, and a third to remove the gene encoding PD-1 that identifies the T cells as immune cells and can be used by cancer cells to disable T cells. ²²

It is not yet known whether cas, a bacterial protein, will trigger the patient's innate immune system. Another source of concern is off-target effects in the T cells and their potential impact and the low rate of cells in which all three edits will have taken place.

Vector Technology: Viruses Versus Non-Viral Vectors

Gene therapy needs to deliver genes to and into a target cell without degradation. Viral vectors have the advantage that they are evolutionary optimized for exactly this process. However, viral vectors can pose a number of problems. Cell and tissue specificity differs so that viruses might transfect healthy cells alongside cancer cells. DNA-based viruses carry the risk of insertion in genomic locations that might lead to detrimental mutations. This phenomenon has been observed, for example in clinical trials for X-linked severe combined immunodeficiency patients, in which hematopoietic stem cells were transduced with a corrective transgene using a retrovirus. This led to the development of T-cell leukemia in 4/20 patients. Another problem with viruses is the potential induction of an acute and/or lasting immune reaction. Furthermore, if the vector is based on a common virus, a patient may already have acquired immunity against it.

Non-viral vectors on the other hand are much less immunogenic and therefore cause fewer safety problems. The difficulties, however, are in production, formulation, and storage, as well as in administration, targeting, and intracellular delivery. Researchers have intensively tested physical delivery methods as well as liposome and polymer formulations, and numerous new compounds are under investigation, including chemical modifications of naked nucleic acid vectors. ²³

OVs

OVs have been explored as cancer therapeutics since the 1960s. They demonstrate selective replication within tumor cells and can trigger antitumor immune responses so that they destroy tumor cells either directly or via induction of immunogenic cell death. Moreover, they activate a long-term antitumor response and can prevent cancer recurrence. As of 2016, the FDA had granted five OVs orphan designation. A genetically engineered live oncolytic herpes virus (T-VEC Talimogene laherparepvec) is already marketed in the United States and Europe for the treatment of melanoma lesions in the skin and lymph nodes. The virus has been modified by removing two genes for improved safety in order to ensure tumor-selective replication and to induce host immunity by adding the gene for the granulocyte-macrophage colony-stimulating factor (GM-CSF). ²⁴ T-VEC resulted in increased durable responses, induced antitumor T-cell immunity, and had a favorable safety profile. However, it has not been shown that T-VEC improves overall survival or has an effect on visceral metastases, and combination treatments with immune checkpoint inhibitors ipilimumab and pembrolizumab are currently being tested. ²⁵

There are many open questions, for example about optimum virus type for a particular indication, potential synergistic effects of adding further viruses or treatments, the need for adjuvants, the route of administration, dosage, and so on. OVs are a field of many innovations with vectors, ²⁶ for example the armament of oncolytic adenovirus with immune active cytokines ²⁷ or the introduction of miRNA response elements to improve the selectivity of OVs for tumor cells and to minimize side effects caused by infection of healthy tissue. ²⁸ Oncolytic non-naturally occurring adenoviruses (e.g., EnAdenotucirev [EnAd]) have been shown to express CD80 and CD3, both molecules improving the antigen presentation in tumor cells and thus the overall tumor immunogenicity. ^29,30 Another version of this virus expresses an immune cell engaging bispecific antibody. ³¹

Another example is novel oncolytic viruses that carry additional genes encoding, for example, for a protein that leads to a fusion of infected cells with neighboring cells and for GM-CSF, anti-CTLA-4 antibodies, or other proteins co-stimulating important immune mechanisms. Researchers also hope that the infection leads to immunogenic cell death and the subsequent release of neo-antigens, further stimulating an immune response. Intratumoral administration aims to improve safety by avoiding an immune response to the virus and also allows for a higher concentration at the tumor site. A first clinical trial sponsored by Replimune Group Inc., is expected by mid-2018. ³²

Non-Replicating Viruses

Other examples are VB-111 and Instiladrin. VB-111 (ofranergene obadenovec), a genetically engineered, non-integrating, non-replicating adenovirus serotype 5 vector is being applied systemically. It expresses a promotor and a transgene that serve as an angiogenesis-specific sensor specifically to induce cell death in angiogenic endothelial cells in the tumor milieu. It has demonstrated tissue and condition specificity as well as efficacy. ³³ At present, the approach is being evaluated in a Phase III trial in combination with Bevacizumab in recurrent glioblastoma patients (www.clinicaltrials.gov/ct2/show/NCT02511405).

Instiladrin (rAd-IFN with Syn3), a genetically engineered adenovirus serotype 5 vector expressing IFN alpha 2b is under development for the treatment of bladder cancer (www.clinicaltrials.gov/ct2/show/NCT02773849).

Genetically Engineered Non-Viral Vectors

Rather than using viruses that may raise issues in terms of immunogenicity and safety, scientists have explored a host of novel technologies to deliver nucleic acids either locally or systemically to target tumor tissue. Vectors may consist of DNA or RNA and are packed in nanoparticles, liposomes, and so on, with or without chemical modification and stabilization.

In April 2010, sipuleucel-T (Provenge^®), the first vaccine for cancer treatment, was approved for use in metastatic prostate cancer. It is designed to stimulate an immune response to prostatic acid phosphatase (PAP), an antigen that is found on most prostate cancer cells. In clinical trials, sipuleucel-T increased the survival of men with a certain type of metastatic prostate cancer by about 4 months. ³⁴

In addition to the limited efficacy, the process of manufacturing the vaccine is complex and costly. To avoid this laborious procedure, researchers tried to develop off-the-shelf vaccines capable of inducing a potent, lasting T-cell response without an ex vivo procedure. Most of these experimental compounds were based on synthetic peptides, some on DNA or RNA vaccines. However, many of these vaccines did not show clinical efficacy.

There have been numerous attempts to tackle this problem by designing vaccines targeting several antigens at the same time, trying to elicit strong T-cell responses against multiple antigens expressed by the same tumor, but still success is limited.

However, the experiences lead to the identification of common success factors for cancer vaccines: an effective route of administration, the generation of sufficient antigen concentration in DCs, the use of an adjuvant that activates DCs, the choice of combination therapies, and, most importantly, the selection of the right antigens and the use of several of those antigens in one vaccine. The last two observations have led to the conclusion that it is best to identify, select, and combine so-called neo-antigens, that is, antigens that are not shared between cancer and healthy cells, thereby avoiding central immunological tolerance. It has also become clear that cancer vaccines need adjuvants and/or combination with other agents to combat immune evasion strategies.

The latest innovation in this field has therefore been personalized vaccines in which researchers combined antigens identified from a patient's tumor sample. The process again is complex and requires a tumor sample and a sequencing step followed by analysis to identify neo-antigens and to select from those the antigens that are most likely to identify a strong and lasting immune reaction.

Results again are mixed. Most of the first wave of cancer vaccines did not succeed in clinical trials for lack of efficacy. Latest results with more advanced cancer vaccines—whether based on nucleic acids or peptides—demonstrate a similar picture. In 2016, a Phase III study of a cancer a vaccine consisting of 10 tumor-associated peptides in patients with metastatic renal cell carcinoma in combination with sunitinib did not meet its primary endpoint of showing an overall survival benefit. ³⁵

In 2017, a Phase IIb clinical trial with a cancer vaccine containing self-adjuvanted mRNA encoding for the self-antigens PSA, PSCA, PSMA, and STEAP1, which are overexpressed in prostate cancer patients, failed to meet the primary endpoint of improving overall survival. Progression-free survival was similar in both arms of the clinical trial. ³⁶

On the other hand, in July 2017, two papers strongly indicated that patients with late-stage melanoma experienced clinical benefit following administration of a personalized cancer vaccine, of which one was based on peptides and the other one on mRNA. ^37,38

However, there are still caveats, as melanoma is characterized by hundreds if not thousands of mutations, so that there is a wealth of potential neo-antigens from which to choose. This is different in other cancers. Furthermore, the process of designing and manufacturing a personalized vaccine—whether based on peptides or on nucleic acids—is complex and time-consuming, taking several weeks to months. As of February 2017, there were at least nine RNA vaccines directed against cancer in clinical trials. ³⁹

The growing understanding of cellular circuits and pathways that make up the tumor-supporting immune system helps the reasons for the mixed results of cancer immunotherapies directed against a single mechanism or single target molecules to be understood. Gene therapy technology enables new approaches in which vectors are designed that deliver gene combinations for the expression of proteins directed against different targets and mechanisms.

The TriMix technology is using three naked mRNA molecules to enhance the activation and maturation of DCs and to activate helper T cells and cytotoxic T cells. The technology uses a synthetic mRNA encoding a CD40 ligand, a constitutively active Toll-like receptor 4 and CD70, together with mRNA encoding fusion proteins of a human leukocyte antigen (HLA)-class II targeting signal (DC-LAMP) and a melanoma-associated antigen (MAA). ⁴⁰ The mix was originally developed to engineer DCs extracorporally, but it is also being explored for intratumoral or systemic delivery. ⁴¹

Another approach uses Immunalon^®, an adenovirus serotype 5 vector expressing three potent immune-modifying signaling proteins—IL-12, IL-2, and CD137 ligand (4-1BBL)—to turn off tumor-mediated immune suppression and to promote T-cell infiltration of tumors and recognition of cancer cells. Instead of addressing one specific immune cell type, these molecules stimulate the activation, proliferation, and survival of a synergizing network of tumor-defensive immune cells in both the innate and the adaptive immune system. ⁴²

Rna-Based Medicine: Easy, Fast, and Cost-Efficient

The growing body of experience with RNA-based medicines—vaccines as well as gene therapies—is outlining an interesting perspective for regulation and manufacturing. For one, from a regulatory standpoint RNA comes without many of the concerns raised with the application of DNA or viruses. There is no risk of unintended or deleterious integration into the genome or the germline, as RNA does not integrate into the genome and is degraded within days, even when stabilized. As a result, effects are transitory and cease once application is halted.

Second, the transfer of RNA sidesteps the tedious manufacturing step for peptide-/protein-based therapeutics. RNA therapeutics can be manufactured chemically or in cell-free systems, without the need for fermenters, concentration, and purification steps and can be formulated to be stored without refrigeration. Recently, a research team demonstrated in an animal model of cancer that it is even possible to deliver RNA systemically so that the animal's liver produces therapeutically efficient concentrations of a bispecific antibody, a quite complex molecule. The bispecific antibody encoded for by the RNA was produced for about a week in sufficient quantities. ⁴³ In the long run, this development might displace many recombinant protein therapeutics, as it is faster, less complex, and more cost-effective.

Conclusion

The most interesting aspect is the versatility of the gene therapy approach. Genetically engineered vectors can be used to transfect all sorts of cells and tissues, ex vivo as well as in vivo, and they can address the most pressing problems in today's cancer therapy: the departure from the one drug–one target paradigm, but also the paradigmatic organ focus in cancer therapy toward the development and application of therapeutics based on immune markers.

Today, gene therapy is the most elegant and feasible option to address multiple targets with one drug, which in the body can lead to the expression of multiple therapeutic molecules addressing immune cells, tumor evasion mechanisms, and so on at once.

For quite some time, cancer immunotherapy has focused entirely on antigens, and only in recent years has it begun to study the multiple mechanisms employed by cancer cells to deactivate the immune system. The development has been sparked by the discovery of checkpoint inhibitors and the subsequent approval of the first drug targeting those inhibitors, which are present in many cancers, independent of the organ from which they have originated.

Regulators are already recognizing this knowledge. Recently, checkpoint inhibitor Keytruda pembrolizumab was approved for the treatment of certain patients with metastatic non–small cell lung cancer (NSCLC) based on biomarker data. Doctors are required to test for overexpression of PD-L1, a PD-1 receptor ligand that plays a major role in suppressing the immune system in NSCLC. This may mark the beginning of a trend in which biomarkers such as immune markers will become more and more important in treatment strategies.

Gene therapy is well equipped with technologies to address these new paradigms. Already it has become a driver of innovation in the immuno-oncology field, enabling drug combinations, novel target definitions, and target validation. It also has tremendous influence in the associated fields of manufacturing, analytical methods, and logistics of biological therapeutics. As a result, it may experience an unexpected renaissance in the future.