Abstract
Since the approval of insulin as the first recombinant therapeutic protein, the prominence of biologic therapies in drug development has grown significantly. Many modalities beyond traditional biologics are now being developed or explored for various indications with significant unmet medical needs. From early traditional replacement proteins to more recent, highly engineered antibodies, oligonucleotides, fusion proteins, and gene constructs, biologic agents have delivered life-changing therapies, despite often having scientifically and technically challenging development programs. This brief review outlines some of the major biotherapeutic classes and identifies the advantages and challenges with the development of these products.
Introduction
The approval of human recombinant insulin in 1982 began an expansive growth in biotherapeutics. Great successes were achieved and multiple life-altering therapies were developed for indications like cancer (e.g. trastuzumab and campath), rare genetic diseases (e.g. cerezyme, fabrazyme, aldurazyme, myozyme, and mipomersan), and immune disorders (e.g. adalimumab and rituximab). Typically, biologics have a favorable attrition rate, though biosimilars are likely going to make the market highly competitive, as more and more biologic therapeutics reach the end of their patent life. Significant guidance has been released by regulatory agencies in the last two decades to help the rational and scientifically based development of these complex products. The International Conference on Harmonisation S6 guidance, adopted in 1997, was the first document to specifically outline the best practices for nonclinical approaches to biologic drug development. 1 Additional guidance documents have since been adopted with regular addendums to reflect scientific advances. 2,3 There remain challenges in the development of biologics that need to be addressed on a case-by-case basis for each particular therapeutic agent. For instance, species specificity of therapeutic action may limit standard study designs for safety evaluation prior to first-in-human dosing. 4,5 Additionally, delivery options are limited, with the main route of dosing so far being either subcutaneous (SC) or intravenous (IV). The manufacturing of these products is also complex, requiring stringent controls and specifications to ensure safety and efficacy. Finally, the immunogenic potential of these products may lead to safety concerns and/or limited efficacy. 6 Despite these challenges, there has been great success in developing and getting these therapeutic agents to markets.
Traditional biologics and monoclonal antibodies
Traditional biologics represent a relatively straightforward approach to disease modification and include replacement proteins, cytokines, and hormones, usually generated from recombinant DNA technologies. Examples of approved traditional biologics include growth factors, insulin, erythropoietin (EPO), enzymes, interferon, and granulocyte colony-stimulating factor.
Standard nonclinical development programs need to be adjusted for these agents based on the nature of the pharmacological targets and the properties of the molecular constructs; in particular, the use of multiple nonclinical species may be inappropriate to identify safety issues relevant to human beings. Nonclinical safety studies with the human therapeutic drug (clinical candidate) should be performed in the most pharmacologically relevant species with comparable binding and affinity to humans to identify potential toxicity issues relevant to humans.
As the activity of biologic therapeutics is frequently species specific, the selection of appropriate animal models or the use of surrogate molecules is crucial to generating useful data. 7,8 Some animal models (such as genetically engineered rodents or disease models) are sometimes used to address interspecies issues, while surrogate molecules pharmacologically active in animal species or specific in vitro strategies may represent appropriate alternatives. The challenge of developing a traditional biologic agent either through the use of a surrogate molecule or the use of an engineered animal model is not trivial, and this approach is associated with important caveats.
The use of relevant animal models of disease and transgenic animals is becoming an acceptable practice for the safety evaluation of biologics. 9 Animal models may generate safety information that is pharmacologically relevant to the human disease. 10,11 However, when using alternate animal models, it is important to understand the differences and limits of the experimental system and to explain, understand, and characterize these differences adequately to make the data relevant to humans.
Nonclinical safety studies are conducted in normal animals or specially aged animals to support the clinical development of some biologic products, especially for rare, life-threatening diseases, where no alternative therapies exist. 12 –15 Many of these studies, particularly those required to advance into pediatric populations, are often irrelevant based on known mechanism and prior safety history and therefore are limiting the advancement of potentially beneficial therapies into the clinic. Hopefully, future guidance addressing the needs of nonclinical safety programs for rare disease populations will become available.
One of the most significant challenges associated with the development of therapeutic proteins is the generation of antidrug antibodies (ADAs). While the generation of ADAs in nonclinical species does not suggest a potential for ADA generation in humans, it may enhance the understanding of the consequence of these antibodies that could be relevant to humans. 16 This is especially important when a neutralizing antibody is developed against a lifesaving protein, rendering the administered protein ineffective. 17 Also, the development of ADAs may neutralize the endogenous protein, thereby creating a clinically relevant deficiency that is adverse to the patient. This was seen with recombinant EPO, where ADAs neutralized the endogenous protein in patients. 18 New and novel treatment regimens are being used and developed to mitigate the generation of antibodies to therapeutic proteins, which will result in sustained efficacy in many patients. 19
In the past three to four decades, there has been an increasing interest for targeted therapeutics, thereby promoting the development of monoclonal antibodies (mAbs), particularly for the treatment of cancer and immunological disorders. In the past, mAbs were primarily used as laboratory tools in biochemistry, molecular biology, and diagnostic medicine, but the strategic direction has shifted due to the evolving knowledge of biologics, perceived benefit based on their favorable characteristics (e.g. selectivity, long half-life, and safety), and enhanced technology for production and purification.
The first mAb approved and marketed for human therapeutic use was Muromonab-CD3 (OKT3) in 1986, a murine antibody. However, murine antibodies have significant drawbacks including a propensity to induce immunogenic reactions, ADAs, and decreased half-lives in humans. 20 –23 The advent of newer antibody technologies has led to the creation of chimeric (typically >50% human), humanized (>90% human), and fully human antibodies to improve the probability of success based on optimized properties (better plasma stability and less immunogenic effects). 24 –29 Adalimumab (an anti-tumor necrosis factor mAb) was the first fully humanized antibody approved (rheumatoid arthritis indication) in 2002, and more have since been marketed or are in early and late clinical development.
Developing mAbs from a nonclinical perspective is unique from that of a small molecule, particularly with selecting the appropriate species for toxicological evaluation. As is the case for all biologics, toxicology studies should only be conducted in the relevant species (species that demonstrates pharmacological activity which is comparable to humans). 2,5,8,30 This is often only the nonhuman primate; however, if a second species is relevant (rodent), it should also be evaluated. Minimally, the sequence homology and binding affinities should be compared between species (e.g. rat, mouse, rabbit, and monkey), but other factors for consideration include functional activity as well as receptor and ligand occupancy and kinetics. If no relevant species can be identified, other animal models or the use of surrogate molecules may need to be considered.
The future of biopharmaceuticals looks promising, yet there are still opportunities, particularly through mAb modifications. 31 Depending on the needs of a specific therapeutic indication, engineering the fragment crystallizable domain of an mAb to increase or decrease effector function and/or to further extend the half-life is being utilized to tailor the properties of the antibody. 32 Additional modifications include bispecifics (with multiple, functionally different binding domains to enable interaction with two target antigens) and antibody drug conjugates, which combine an mAb with a highly potent cytoxic chemical (often a cytotoxic agent). Together these approaches open up a new era in mAb development that is promising for advancing new medicines to patients.
Gene therapy
The Food and Drug Administration (FDA) defines gene therapy as “a medical intervention based on modification of the genetic material of living cells.” 33 In 1990, the first FDA-approved gene therapy trial, which used gene-corrected autologous T cells, was conducted in two children afflicted by adenosine deaminase deficiency with limited success but no significant safety concerns. 34 –36 In the following decade, gene therapy was viewed as a highly promising therapy for monogenic diseases and the only plausible option for many lethal orphan diseases. 35 Gene therapy suffered a major setback in 1999 when the death of a patient on a clinical trial for ornithine transcarbamylase deficiency was directly attributed to an immunologic reaction against the viral vector used for gene delivery. 37 Other drawbacks that contributed to a substantial decline in support for gene therapy during the 2000s included multiple reports of insertional mutagenesis leading to clonal transformation, 38,39 immunotoxicity, 40 gene silencing, 41 and a failure of studies performed in nonclinical species to accurately predict human efficacy and safety. 42
Recent years have seen a resurgence in the use of gene therapy strategies to treat not only rare monogenic disease but also chronic and progressive diseases, including cancer, heart failure, and neurodegenerative diseases. 35 To date, cancer accounts for greater than 60% of all clinical trials involving a gene therapy strategy. 36 This rediscovered enthusiasm for gene therapy has been driven largely by substantial advancements in the design of viral and nonviral vectors, the selection of appropriate therapeutic genes, and the development of targeted delivery strategies that effectively regulate the expression of the transgene while restricting adverse effects. 35,43,44 For example, adenoviral modifications have reduced the immunogenicity of the virus, resulting in greater transgene stability and fewer unwanted immunologic side effects. Gammaretroviral vectors have many unique qualities that lend themselves to effective gene therapy; however, their random insertion into the human DNA has continued to create substantial safety concerns regarding the potential for neoplastic transformation. Recent strategies utilizing self-inactivating lentiviruses that reduce the potential for insertional mutagenesis, but maintain many of the desirable qualities of a retrovirus, may provide an attractive alternative. 41,45 The side effects of viral vectors can be avoided through nonviral methods of gene transfer, but despite improved safety profiles, their use has been limited by a low transduction efficiency. 42 Advances in delivery systems such as cationic lipid-based nanoparticles (lipoplexes) are improving this efficiency, and effective nonviral delivery will likely become a reality in future studies. 43
Over the past 50 years, the field of gene therapy has experienced great highs and lows. The theoretical simplicity as to how this powerful technology could cure disease generated tremendous excitement in the scientific community and the public domain. Results of initial clinical trials, however, were sobering and what logically followed were several decades of careful research and the development of substantially safer and more effective methods. Recent successes in clinical trials have been achieved for a number of conditions including heart failure, 46 B cell malignancies, 47 leukodystrophy, 48 β-thalassemia, 49 hereditary immunodeficiency diseases, 38,39,41 Leber’s amaursosis, 50 Parkinson’s disease, 45 and hemophilia B. 51 Europe officially approved the first marketed gene therapy product (Glybera®) for lipoprotein lipase deficiency in 2012. Given these positive developments and the realistic approaches that were conceived following a series of dramatic setbacks, gene therapy is poised to become an important therapeutic modality for unmet medical needs. 4
Nucleic acid therapeutics
Nucleic acid therapeutics (NATs) comprise various agents that aim at inhibiting the expression of specific target molecules (e.g. messenger RNAs (mRNAs), microRNAs (miRNAs), viral transcripts, and proteins). These agents include antisense oligonucleotides (ASOs), nucleic acids acting through the RNA interference pathway (e.g. small interfering RNAs (siRNAs) and miRNAs), ribozymes (catalytically active molecules that cleave target RNAs), and aptamers (which target proteins). Additional types of NATs will likely emerge in the future fueled by innovation. NATs are typically administered IV or SC, but local injections (e.g. intravitreal and intralesional) and inhalation have also been used. 52 –54
ASOs were the first NATs investigated nonclinically and clinically. These short (approximately 18–25 nucleotide long) single-stranded oligonucleotides bind specific mRNA sequences through Watson–Crick base pairing, thereby resulting in RNase H-driven cleavage or translational block of the targeted transcript. 55 ASOs have been in development for more than two decades for multiple indications, but only two products have been approved by the FDA: fomivirsen (vitravene) for the treatment of cytomegalovirus retinitis 56 and kynamro (mipomersen) for the treatment of homozygous familial hypercholesterolemia (it is noteworthy that the Committee for Medicinal Products for Human Use of the European Medicines Agency has recommended the refusal of the marketing authorization for kynamro. ). 57,58 Most of the discussion below focuses on ASOs because of the abundant experience with these agents, but most findings are relevant to the other types of NATs.
The lack of approved NATs in spite of considerable investments 59 reflects the challenges associated with these agents (the keywords ‘antisense’ and ‘siRNA’ resulted in a list of 126 and 37 ongoing or completed clinical trials, respectively). 55,60 It is beyond the scope of this review to cover in detail all these challenges, and readers are referred to several recent reviews and commentaries. 55,61,62 Briefly, the major challenges associated with the development of NATs include: (1) instability related to nuclease degradation in plasma and tissues, (2) delivery to the desired tissue(s), (3) potency to silence the target of interest, (4) off-target effects due to partial sequence complementarity, and (5) sequence-dependent, but hybridization-independent toxicity. Stability has been addressed through various chemical modifications, while progress in improving potency and limiting off-target effects has been achieved through the use of computational methods and microarray-based screens. 55,62 Delivery of NATs still remains, probably, the most challenging hurdle to address and represents an active field of investigations. 53,69
The toxicity issues associated with NATs and ASOs in particular have been recently reviewed. 63 Reversible, hybridization-independent pro-inflammatory effects can be observed due to recognition by receptors sensitive to exogenous nucleic acids, such as the Toll-like receptors. 63 This activation of the innate immune system manifests in the clinic as flu-like symptoms. Interestingly, this property has been utilized as an immunostimulatory approach for cancer therapy and vaccine adjuvants. 60 In addition, hepatotoxicity, kidney toxicity, thrombocytopenia, complement activation (especially in monkeys), and inhibition of coagulation have been reported in nonclinical species and/or humans with ASOs. 63 –66
No specific regulatory guidance is available for the development of NATs. Given their structure, it is reasonable to consider them like small molecules and apply standard nonclinical approaches, such as the use of both a rodent and a non-rodent species. Since species difference may exist in the sequence of the target(s) of interest, the use of rodent NAT surrogates may be warranted. Sufficiently sensitive bioanalytical methods are necessary for pharmacokinetic evaluation. 67 Finally, ASOs have not demonstrated the effects in dedicated safety pharmacology (neurobehavior, respiratory, and cardiovascular) studies, suggesting that these studies are not warranted for ASOs. 68
Advances in delivery systems (especially those enabling targeted tissue delivery) and in RNA biogenesis will likely translate into some clinical successes in the near future. Indeed, there appears to be a regained interest in NATs by the industry. However, while these advances will address the current development barriers, new challenges will arise. For example, the current delivery systems result in accumulation of ASOs mostly in the liver, kidney, and tissue macrophages. This biodistribution will be affected with novel biomolecular delivery systems. Likewise, the safety profile of these novel delivery systems is for the most part unexplored. It will take definitely significant efforts by the research community before the clinical utility of NATs becomes a reality.
Conclusion
Biologic agents have already shown clinical benefit, and as the field keeps expanding, fueled by innovation in molecular engineering and delivery systems, new and exciting therapeutics will be delivered by the pharmaceutical industry with a significant impact on human health. As these agents increase in complexity, nonclinical safety scientists will face new challenges to support the development of these innovative products and will uncover novel, often unexpected safety issues, which will require significant efforts to understand and monitor. As the past has taught us, not all safety issues observed nonclinically are relevant to humans. Likewise, not all human safety issues can be identified nonclinically, especially with poorly designed nonclinical studies or irrelevant animal test systems. Advancing these novel therapeutics safely in the clinic requires not only a strong understanding of the biology of the target but also a good comprehension of the caveats so that limits of the current nonclinical models and an ability to design fit-for-purpose, creative nonclinical safety testing funnels adapted to the test agents are being developed.
Footnotes
Conflict of interest
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
