Role of Drug Repurposing in Sustainable Drug Discovery

Abstract

The contemporary global drug discovery scenario, in spite of several technological advances, is heavily ridden with multiple challenges of a dynamic regulatory system, escalating costs from bench to bedside investigational drugs, the increased probability of withdrawal after launch, and over-stretched timelines from discovery to approval, among others. Drug repurposing/repositioning/re-profiling/re-tasking is an effective and practical complimentary method for the selection of alternate therapies for approved, shelved, discontinued/abandoned, and investigational drugs or new chemical entities, with the parallel study of new metabolic pathways and/or protein targets. Such an approach encompasses multipronged benefits of redundant preclinical testing, toxicity evaluation, and formulation studies, based largely on serendipity. In recent years, approaches have been driven by artificial intelligence (AI) and machine learning, and bioinformatics have opened up new vistas in drug re-profiling acceleration. Increasing protocols to club the shared mechanisms among structurally diverse/dissimilar drugs include pathway analysis, phenotypic screening, signature matching, related disease genes, binding assay studies, molecular docking, and clinical data monitoring. All in all, repositioning of abandoned/investigational/existing drugs or new chemical entities for other therapeutic indications could enhance the overall productivity of the pharmaceutical industry while paradigmatically shifting the focus from new drug discovery to the optimization of available resources.

1. Introduction

The overall global population exposed to regular medicines has increased twofold over the past few decades. To complement this, the average life expectancy has increased considerably, and newer, lesser understood diseases are on the rise. Unfortunately, the rather long timelines of drug design and approval (10-12 years), increased rates of USFDA failure/recall after launch, escalating resources for new drug discovery and development, and a paradigm shift toward green chemistry have, in totality, rendered the conventional drug discovery process largely wanting for alternative backup plans.

According to the Brundtland Commission of the United Nations: “Sustainable development is development that meets the needs of the present without compromising the ability of future generations to meet their own needs” [1]. The Sustainable Development Goals (SDGs) and the One-Health approach are analogous initiatives [2]. The various alternative ways of making drug discovery sustainable are: signature mapping, pathway matching, in silico screening and molecular docking, genetic association, retrospective analysis of clinical data, drug repurposing, high-throughput screening, and so on.

Out of all the options available for introducing sustainability in drug discovery and development, drug repositioning, also known by its alternative names of repurposing, re-profiling, or re-tasking, is the strategy of choice as the advantages far outweigh the challenges encountered in offering a drug for a new medical indication, totally distinct from its original scope.

The most important reason for the failure/withdrawal of an approved drug is addressed adequately: the potentially repurposed drug would have been in preclinical models and early-stage human clinical trials (phase I & II), thereby justifying its safety; thus, subsequent efficacy trials would be more predictable.

In addition, drug development timelines can be sufficiently squeezed, due to non-repetition of preclinical testing, safety assessment, and/or formulation development. Most significantly, it is the most economically lucrative of all the sustainable strategies employed, as mandatory investment is marginal. The regulatory and phase III costs can involve substantial savings in preclinical, phase I, and phase II expenditure.

Since the commencement of drug repurposing concept, it was rational but serendipitous; that is, an approved drug or one in clinical trials was studied for its contraindications, off-target reactions, and/or an enhanced on-target response, patented and groomed for repurposed launch. Most success stories of relaunch in a new therapeutic avatar have indeed relied hugely on serendipity, rather than on a structured and well-planned approach. Such examples include aspirin, minoxidil, sildenafil, thalidomide, celecoxib, rituximab, raloxifene, fingolimod, dapoxetine, topiramate, and ketoconazole, among others.

However, with several technological advancements, in the form of various approaches, drug repurposing in general, and specifically for rare diseases, comprising a databank of over several thousand (~7000), with 95-96% of these having no approved therapeutic agent, primarily because of unknown disease pathophysiology, contemporary drug repurposing abounds in opportunities to fill up the unmet medical space.

As we shall see further, the trade-off between challenges posed by drug repurposing and its many advantages is also laced with limitations of collation, integration, analysis, and interpretation of big data, of biomedical, clinical, pharmacological, or sequencing type.

2. Types of sustainable approaches to drug discovery

The many approaches toward sustainable drug discovery are listed below (Figure 1):

Figure 1.

Sustainable drug repurposing approaches.

2.1 Signature mapping

In a broad sense, this process is based on the relative comparison of certain distinct features, thus referred to as the ‘signature’ of one drug against another, disease pathway comparison, or clinical phenotype matching [3, 4]. This signature pattern of two comparable drugs is generally assigned from three databanks:

1. Transcriptomic (RNA), proteomic, or metabolomic data

2. Chemical structures

3. Adverse event profiles

For example, two drugs were identified using drug–disease similarity approach using the correlation between the gene expression signature of the drug and that of the disease, prednisolone, and topiramate.

2.2 Pathway/network matching

Genetic, protein, or disease data can aid the identification of repurposing targets.

In some disease pathways, the relevant genes are not druggable as targets. This is when a pathway-based comparison of genes that are either downstream or upstream or a genome-wide association studies (GWAS)-related target may be used for repurposing issues [5]. Network analysis uses available information on gene expression patterns, disease pathology, protein interactions, or GWAS data for construction of a drug or disease network to further potential repurposing candidates.

For example, pranlukast, an asthma drug/leukotriene receptor 1 antagonist, and a phosphodiesterase inhibitor amrinone are used for congestive heart failure.

2.3 High-throughput screening

Or phenotypic screening of compounds, it is applied to in vitro or in vivo disease models can reveal potential drug candidates for clinical evaluation.

For example, the discovery of disulfiram, used against alcohol abuse, turned out to be a selective antineoplastic agent, with proven research using genome-wide gene expression studies.

2.4 In Silico screening and molecular docking

It is a computational approach based on structure similarity to predict binding site complementarity between a ligand (for example, a drug) and a therapeutic target (typically, a protein receptor).

For example, molecular docking studies identified mebendazole, an anti-parasitic drug and inhibitor of vascular endothelial growth factor receptor 2 (VEGFR2), and also a mediator of angiogenesis. This was validated with experiment studies.

2.5 Genetic association

Genes that assumed to be associated with a disease mechanism should be considered as potential drug targets about which a small molecule ligand can be designed.

2.6 Retrospective analysis of clinical data

Large volumes of old, retrievable clinical data including hospital records, clinical trial data, and post-marketing surveillance data can be tapped, compiled, integrated, and analyzed. Such databases comprise both structured and unstructured data on patient response and outcomes.

For example, terbutaline sulfate, an anti-asthmatic drug candidate, was arrived at for the treatment of amyotrophic lateral sclerosis (ALS), from the retrospective clinical data analysis.

3. Drug repurposing

3.1 Challenges to overcome

3.1.1 Legal and commercial challenges

Some developing countries like India, the Philippines, Taiwan (Province of China), the Andean Pact Latin American countries (Bolivia, Colombia, Ecuador, and Peru), and Vietnam do not allow the granting of a method-of-use patent for a second or alternate medical use of an approved drug, whereas several developed nations allow second medical use patents defined as fiction of novelty. In some other cases, possible repurposing may have been discussed/covered/reported in the original literature or may be part of its non-registered uses. The international TRIPS Agreement allows flexibility to the individual signatory nations as also to their courts on the refusal/approval of second medical use patents.

The European Patent Office (EPO) does not allow a second medical use patent if it is merely noted that the drug exhibits selective binding to another receptor; yet if the claim focuses on the end result of the drug function, mentioned categorically as “any condition susceptible of being improved or prevented by inhibition or enhancement of a specific enzyme activity,” the patent is granted/approved provided that it is validated/supported by experimental data, which is also disclosed in the patent application specifications. Any information related to drug re-profiling may not be controlled by intellectual property rights; however, if in the public domain, it cannot attract novelty, thus ceasing to be patentable material. Furthermore, proof-of-concept studies also need to be validated by controlled clinical trials.

Toggling the bioisosteres of existing off-patent drugs by keeping their pharmacophore intact would defeat the purpose of repurposing, as it would give rise to a new entity. Same dosage and/or formulation for off-patent drugs being re-profiled would not be legally patentable. Therefore, offering a lower dosage form than in use or an alternate formulation would be key to its re-profiling.

3.1.2 Retrieving data

With respect to shelved/abandoned/withdrawn drugs, a major hurdle is accessibility to clinical trials data that only its discovering organization may be privy to. Although a shelved drug is capable of reinventing itself as a lower dose repurposed opportunity, the pharma industry, on an average, does not resort to sharing its list/portfolio of shelved drugs, especially if the therapy area of repurposing does not cater to its organization’s disease portfolio.

In another scenario, where there is a collaborative interface between industry and academia, this problem can only be facilitated by CDA (Confidentiality Disclosure Agreement) and compound sharing. Dealing with abandoned drugs, or those which have been outlived by their competitors, another challenge is posed by the availability of a suitable vendor.

3.1.3 Limited repurposing space

As more and more approved drugs are finding their way into repurposed therapeutic territories, it may seem that the druggable space for repurposing is exhausting itself out. In such a scenario, a prudent strategy would be to club drugs for combination therapy, as also to discover new pathways for their effective applications. This is visible more so in the field of infectious diseases, by employing the nifurtimox–eflornithine combination therapy for second-stage African trypanosomiasis [6]. Another avenue to expand the re-profiling horizon is personalized medicine.

4. Early examples of successful drug repositioning

The first example of drug repositioning is aspirin or acetylsalicylic acid (Figure 2). In 1899, Bayer marketed it as an analgesic, while at the turn of the century, in the 1980s, it was repositioned as an antiplatelet aggregation drug [7].

Figure 2.

Aspirin/AcetylSalicylic acid.

During the last millennium, drug repurposing successes of sildenafil, valproic acid, and minoxidil were charted by analyzing their known pharmacology in a particular domain (vis-a-vis the adverse effects) to resolve a pressing clinical problem from another therapeutic area, not relying entirely on serendipity [8].

Sildenafil (formulated as its citrate) (Figure 3), originally introduced by Pfizer, as an antihypertensive medication, was later repurposed for erectile dysfunction (TM Viagra) based on prior clinical experience and went on to become a blockbuster drug [9]. Moreover, as a reversible inhibitor of phosphodiesterase type 5 (PDE5), it has also been approved for pulmonary arterial hypertension (PAH) intervention under the brand name of Revatio® [10].

Figure 3.

Sildenafil citrate.

Valproic acid (N-dipropylacetic acid) (Figure 4) was discovered by Meunier and Carraz in the year 1967. On the revelation of its anticonvulsant properties, it was a popular drug widely used in epilepsy and bipolar disorder, formulated as sodium valproate. Valproic acid (VPA) has so far been the drug of choice for epilepsy and other neurological disorders since the past 66 years. Ongoing research has indicated the potential of VPA as an antineoplastic agent, partly due to its role in the inhibition of histone deacetylases, thus modulating the expression of genes and affecting changes in the cell cycle, differentiation, and subsequently apoptosis. Over and above inhibiting histone deacetylases, VPA enhances RNA interference, activating histone methyl-transferases, or suppressing the activation of transcription factors [11].

Figure 4.

Sodium valproate.

Minoxidil (Figure 5) was originally developed in the late 1950s, by The Upjohn Company, now Pfizer, with hopes of treating ulcers; however, it failed to treat gastric issues, and instead was shown to be a vasodilator and a potassium channel opener, hyperpolarizing the cell membranes. Two decades later, in 1979, minoxidil was repurposed for arterial hypertension [12]. During clinical trials, it was observed that unwanted hair growth was an adverse side effect, and thus, subsequently, a topical minoxidil formulation was evaluated to treat hair loss. FDA approved topical minoxidil in 1988 for androgenetic alopecia and alopecia areata, and today, generic minoxidil is sold over the counter (OTC) as Regaine Topical Solution 2%, for men and women [13].

Figure 5.

Minoxidil.

Another drug, thalidomide, originally introduced as a sedative in the year 1957, was found to induce severe skeletal birth defects in newborn children whose mothers were administered this drug in the first pregnancy trimester; thus, after about 4 years, it was withdrawn and further repurposed for erythema nodosum leprosum (ENL) (1964), and multiple myeloma (1999) was based on serendipity [14]. The two indications were distinct of each other, and decades apart from each other. The successful repurposing of thalidomide also led to the discovery, development, and approval of other highly successful derivatives, notably lenalidomide (Revlimid, Celgene) (Figure 6).

Figure 6.

Thalidomide and its derivatives, Lenalidomide, Pomalidomide, and BTX306.

This strategy is especially very pertinent when applied to rare and neglected diseases and disorders, and orphan drugs, as is noticed in the drug portfolio of DNDi (Drug for Neglected Diseases initiative) repurposed NCE candidates undergoing clinical trials, including fexinidazole, fosravuconazole, Ambisome™, and miltefosine, fexinidazole being the first oral-only drug, for advanced-stage sleeping sickness [15], with a small fraction of the estimated investment for a de novo drug. Rare and orphan diseases often have unknown or poorly characterized metabolic pathways/pathophysiology where computational approaches for predictive repurposing result in large-scale genome sequencing data analysis for the identification of genetic variation/s contributing to the disorder and expediting the re-profiling of existing drugs targeting the relevant protein/s [16].

During the past decade, drug repurposing as well as drug discovery is hugely complemented by artificial intelligence/machine learning methods to squeeze the drug discovery timelines while maintaining consistency in the systematic retrieval, analysis, and application of big data. Notable approaches in this area are systematic analysis of clinical trial data, molecular similarity approximations, signature matching of targets, transcriptomic and proteomic data, and structure-based virtual screens.

5. Unusual case studies

5.1 Dapoxetine

Dapoxetine, discovered by Eli Lilly, as a selective serotonin reuptake inhibitor (SSRI), was primarily discontinued by Eli Lilly as an adjunct therapy for analgesia. Later on, it was repurposed as an antidepressant with fluoxetine. But short half-life of the compound and the rapid onset did not permit daily dosage, an imperative criterion for any supplementary antidepressant, and thus, it fizzled out.

Finally, dapoxetine’s rapid onset and short half-life were considered to be a pharmacokinetic advantage for the therapy of premature ejaculation, which prompted patenting the findings and validating with phase II proof-of-concept studies, and after changing hands, with a new method-of-use patent, dapoxetine (now a part of Johnson & Johnson portfolio) is now in phase III clinical development for premature ejaculation.

5.2 Thalidomide

As noted previously, after severely unfortunate fallouts, thalidomide made a grand role reversal. Originally marketed as a sedative in 1957 in Germany and England, Thalidomide created unforeseen complications in pregnant women facing morning sickness. It was subsequently repurposed as a drug of choice to treat the erythema nodosum laprosum (ENL), an agonizing inflammation caused due to leprosy whereby painfully large boils often lead to blindness.

6. Future outlook

Currently, almost a third of approvals are repurposed drugs. This clearly indicates the success of several repurposed drugs. New avenues for repositioning can emerge from increased collaboration between pharmaceutical industry and allied sectors of academia, with priority awarded to orphan diseases, rare and neglected disorders, and synergistic drug combinations of repurposed drugs, in such cases as metformin, where the efficacy of the original drug continues unabated. Rare and neglected diseases are not usually profitable for pharmaceutical giants, and their involvement in terms of corporate social responsibility, which endeavors to bridge the gap between profit margins and overall societal welfare, can also add up to generate awareness about several lesser-known disorders. These can be heavily incentivized, parallelly by governmental agencies/organizations, to sustain the equilibrium between commercial viability and redressal of new therapeutic solutions for marginalized diseases.

To supplement this, personalized/precision medicine will add to newer information regarding the characterization and classification of disease pathways, leading to a better understanding of repositioning old/abandoned drugs for diverse therapeutic areas, revitalizing drug re-profiling even further. Contraindications/adverse reactions relating to a prior allotted pharmacokinetic metabolism can assist and enhance our knowledge of expected drug reactions. Sustainable drug re-profiling is also complimented by advances in technology such as artificial intelligence and machine learning, which are instrumental in the expedition of large-scale extraction and integration of heterogeneous data, comprising a mixed bag of imaging, HTS data, DMPK profiles, clinical trials records, and electronic health reports and records, to name a few.

Thus, de novo drug discovery and drug repositioning can support each other to make pharmaceutical innovation more sustainable in terms of resources, time, and setbacks.